Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Characterizing subgroups of sexual behaviors among men who have sex with men eligible for, but not using, PrEP in the Netherlands

  • Feline de la Court ,

    Roles Conceptualization, Formal analysis, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing

    fdlcourt@ggd.amsterdam.nl

    Affiliation Department of Infectious Diseases, Public Health Service of Amsterdam, Amsterdam, The Netherlands

  • Daphne van Wees,

    Roles Conceptualization, Methodology, Supervision, Writing – review & editing

    Affiliation Center for Infectious Disease Control, National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands

  • Birgit van Benthem,

    Roles Writing – review & editing

    Affiliation Center for Infectious Disease Control, National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands

  • Elske Hoornenborg,

    Roles Writing – review & editing

    Affiliation Department of Infectious Diseases, Public Health Service of Amsterdam, Amsterdam, The Netherlands

  • Maria Prins,

    Roles Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing

    Affiliations Department of Infectious Diseases, Public Health Service of Amsterdam, Amsterdam, The Netherlands, Department of Infectious Diseases, Amsterdam Institute for Infection & Immunity (AII), Amsterdam UMC, University of Amsterdam, Amsterdam, The Netherlands

  • Anders Boyd

    Roles Conceptualization, Methodology, Supervision, Writing – review & editing

    Affiliations Department of Infectious Diseases, Public Health Service of Amsterdam, Amsterdam, The Netherlands, Department of Infectious Diseases, Amsterdam Institute for Infection & Immunity (AII), Amsterdam UMC, University of Amsterdam, Amsterdam, The Netherlands, Stichting HIV Monitoring, Amsterdam, The Netherlands

Abstract

This study identified subgroups of sexual behaviors associated with increased STI/HIV risk among those eligible for but not using pre-exposure prophylaxis (PrEP) in order to improve PrEP uptake and prioritization in the context of restricted capacity. We used data from sexual health centers (SHCs) in the Netherlands, including all visits of eligible but non-PrEP using men who have sex with men (MSM), men who have sex with men and women (MSMW) and transgender persons between July 2019 (start of the Dutch national PrEP pilot (NPP)) and June 2021. Using latent class analysis (LCA), we identified classes of sexual behaviors (number of partners, chemsex, group sex and sex work) and explored whether these classes were associated with STI diagnosis and sociodemographics. Across 45,582 visits of 14,588 eligible non-PrEP using individuals, the best fitting LCA model contained three classes of sexual behaviors. Classes were distinguished by seldomly reported sexual behaviors (class 1; 53.5%, n = 24,383), the highest proportions of ≥6 partners and group sex (class 2; 29.8%, n = 13,596), and the highest proportions of chemsex and sex work (class 3; 16.7% of visits, n = 7,603). Visits in classes 2 and 3 (vs. class 1) were significantly more often with individuals who were diagnosed with an STI, older (≥36 vs. ≤35 years), MSMW (vs. MSM), and visiting an urban (vs. non-urban) SHC; while these visits were significantly less often with individuals from an STI/HIV endemic area. The percentage of visits at which an STI was diagnosed was 17.07% (n = 4,163) in class 1, 19.53% (n = 2,655) in class 2 and 25.25% (n = 1,920) in class 3. The highest risk of STI, and thereby HIV, was in those engaging in specific subgroups of sexual behavior characterized by frequently reporting multiple partners, group sex, sex work or chemsex. PrEP uptake should be encouraged and prioritized for these individuals.

Introduction

Pre-exposure prophylaxis (PrEP) is a highly efficacious biomedical HIV prevention strategy [1]. In the Netherlands, PrEP implementation is targeted towards individuals with an increased risk of acquiring HIV, including men who have sex with men (MSM), men who have sex with men and women (MSMW) and transgender persons (TGP) [1, 2]. The current Dutch PrEP eligibility criteria pertain to MSM or TGP who have had (1) condomless insertive and/or receptive anal sex with a male partner with unknown HIV status, or with a known HIV-positive partner with a detectable viral load, (2) an anal sexually transmitted infection (STI), (3) syphilis, or (4) used post-exposure prophylaxis (PEP) in the past 6 months [2]. In 2019, a five-year national PrEP pilot (NPP) was implemented at sexual health centers (SHCs) in the Netherlands, offering subsidized PrEP to 8,500 eligible users [3]. However, because the provision capacity of the NPP is restricted [3], it remains unclear whether there are deficiencies in PrEP uptake among certain subgroups of eligible individuals. Identifying and characterizing eligible non-PrEP users with the highest PrEP need is thus crucial.

In addition to the Dutch PrEP eligibility criteria, certain sexual behaviors are associated with an increased risk of engaging in condomless anal sex and acquiring an HIV infection, which include having multiple partners [4], chemsex [5], group sex [6], and sex work [7, 8]. Previous studies have shown that chemsex [9] and sex work [10] are also associated with lower PrEP uptake; as are certain sociodemographic characteristics, such as younger age (i.e., <30 or <35), living in a non-urban area, lower education level (i.e., no post-secondary education), or having a migration background from an HIV/STI endemic area [1113]. Such behaviors and characteristics can help more effectively recognize PrEP need and prioritize individuals for the NPP beyond the current eligibility criteria. This study therefore aimed to identify subgroups of sexual behavior among MSM, MSMW and TGP who were eligible for, but did not use, PrEP between July 2019 and June 2021. Furthermore, we aimed to explore whether these subgroups were associated with STI diagnosis and sociodemographic variables. Considering that many of the sexual behaviors are highly correlated, making it difficult to construct models examining independent pathways to these outcomes, we used an approach that clustered like behaviors within individuals as latent classes to accomplish these study aims.

Methods

Study design

Data from the Dutch national STI and HIV surveillance database were used. Data are collected from a nationwide system of 24 public sexual health centers (SHCs) across 9 regions in the Netherlands [14]. Free-of-charge STI/HIV testing and care is offered at these centers, which are targeted towards populations at higher risk for STI or HIV infection [14]. Collected data pertain to STI and HIV testing and diagnosis, self-reported sexual behavior, and additional sociodemographic and sexual health-related characteristics. Use of data was requested for analysis in July 2021, which pertains to data collected from July 2019 until June 2021.

All individuals visiting an SHC were of age and provided both verbal informed consent and an opt-out option for sharing data with the RIVM, documented by the SHC. All collected data are coded, secured and fully pseudonymized in accordance with Dutch privacy legislation.

Study population

Between July 2019 and June 2021, we included all visits of HIV-negative MSM, MSMW and TGP with at least one visit at an SHC and who did not report any PrEP use in the preceding 12 months. We further included only visits from individuals who met at least one PrEP eligibility criterion according to the Dutch guidelines [2].

Study variables

Sexual behavior characteristics were the number of partners, chemsex (defined as cocaine, ketamine, mephedrone, gamma-hydroxybutyrate, gamma-butyrolactone, and/or crystal meth use around the time of or during sex [15]), group sex and sex work. Additional variables were STI (defined as anal chlamydia, anal gonorrhea, hepatitis C virus, hepatitis B virus, and/or syphilis diagnosis) and sociodemographic variables, including age (younger/older; cut-off based on median age), education level (low-middle/high; low-middle = primary, secondary, vocational or specialist education, and high = associate, bachelor, master or doctoral degree [16]), sexual partner(s) (male/male and female), originating from an STI/HIV endemic area (yes/no; defined according to the National Institute for Public Health and the Environment (RIVM) as being born in or having either one or both parents born in Suriname, Turkey, Netherlands Antilles, North Africa, Sub-Saharan Africa, Eastern Europe, Central and South America, or Asia [3, 17]) and region of SHC visit (urban/non-urban; where urban is all “Randstad” provinces and non-urban is all other provinces). All variables refer to the six months prior to each visit except for any STI, which pertains to STI diagnosed at the current visit.

Statistical analysis

Descriptive statistics were calculated for sexual behavior and both STI and sociodemographic variables. Latent class analysis (LCA) was performed to distinguish classes of sexual behavior that could be targeted for PrEP uptake. The sexual behaviors examined in LCA were number of partners, chemsex, group sex and sex work in the past six months.

We used a generalized structural equation modelling approach [18]. Briefly, a latent variable model was constructed, based solely on sexual behavior variables, where the probability of engaging in each behavior given latent class k was modelled by an intercept, αdk, specific to each sexual behavior variable d and class k. Sexual behavior characteristics were modeled as dichotomous variables: number of partners (≥6/≤5; cut-off based on median), chemsex (yes/no), group sex (yes/no) and sex work (yes/no). Models were estimated using maximum likelihood, calculated by summing all conditional likelihoods of each latent class multiplied by the associated latent class probabilities. The posteriori probability of a visit i belonging to each class k, πik, was determined from this likelihood. Visits were then assigned a latent class k corresponding to the highest probability πik. We examined participant characteristics of each assigned class across visits. The conditional probabilities of each item (i.e., sexual behaviors) were estimated with intercept-only logit models for each item within classes.

We determined the number of latent classes (1 to 6) using the Bayesian Information Criterion value (BIC) and Akaike Information Criterion (AIC) score, for which lower values indicate a better fit, and an entropy calculation ranging from 0–1, where a higher value indicates higher ability of the model to classify clusters (i.e., degree of class membership separation) [19]. Using the best fitting model, the univariable odds ratios (ORs) and 95% confidence intervals (CIs; using the delta method) for the associations between STI diagnosis or sociodemographic variables (i.e., STI, number of partners, chemsex, group sex, sex work, age, education level, sexual partner(s), originating from an STI/HIV endemic area, and region of SHC visit) and latent classes were obtained from the modeled parameter estimates. Multivariable ORs were then calculated from the parameter estimates of a model including all covariates. Variance estimates were corrected for repeated observations within individuals using a clustered sandwich estimator.

Sensitivity analyses were performed to assess potential bias related to calendar year and to the inclusion of multiple visits per individual. The LCA model was hence repeated (i) for each year, separately, and (ii) using only one randomly selected visit per individual.

For all analyses, STATA (v16, College Station, TX, USA) statistical software was used. Latent class models were estimated using the “gsem” command, and “predict” post-estimation commands were used for posteriori probabilities.

Results

Study population

From the 103,216 visits that took place at SHCs from July 2019 to June 2021 among non-PrEP using MSM who were HIV-negative at first visit in the study period, we excluded 253 visits for the 250 individuals who became HIV positive from the moment of diagnosis, and 57,381 visits at which individuals were not eligible for PrEP. In total, 14,588 individuals were included in analyses, contributing 45,582 visits (median visits per individual = 2; IQR = 1–4).

At first visit between July 2019 to June 2021 (Table 1), the majority was MSM (98.5%), a small proportion was TGP (1.5%), and of the MSM, around 17.7% were MSMW. At this visit, most individuals were aged ≤35 years (59.2%), had a high level of education (59.0%), were not from an STI/HIV endemic area (83.1%), and were visiting an SHC in an urban area (65.7%).

thumbnail
Table 1. Study population characteristics stratified by class membership at first visit between July 2019 and June 2021.

https://doi.org/10.1371/journal.pone.0284056.t001

Classes of sexual behavior

LCA revealed three different classes of sexual behaviors at which individuals were eligible for but did not use PrEP between July 2019 and June 2021. This 3-class LCA model showed the best fit based on higher entropy, and lower BIC and AIC compared to models with more or less classes (S1 Table). Class 1 was characterized by a low number of partners (≤5) and infrequent reports of recent chemsex, group sex and sex work, all in the 6 months prior to the visit (Fig 1). Class 2 was characterized by a large proportion of visits at which a high number of partners (≥6) and group sex were reported. Class 3 was characterized by a high number of visits at which numerous partners (≥6), chemsex and group sex were reported, along with relatively more frequent reports of sex work compared to class 1 and 2.

thumbnail
Fig 1. Sexual behavior associated with increased STI risk across three latent classes.

Explanation of data: Bars represent the mean proportion of visits reporting each sexual behavior respectively for class 1, 2 and 3. All sexual behaviors refer to the six months prior to the visit. Number (= No.) of sexual partners refers to those with ≥6 sexual partners in the six months prior to the visit. Chemsex was defined as using cocaine, ketamine, mephedrone, gamma-hydroxybutyrate (GHB), gamma-butyrolactone (GBL), and/or crystal meth around or during sex. Bands at the top of each bar represent 95% confidence intervals, which were calculated using the delta method.

https://doi.org/10.1371/journal.pone.0284056.g001

The proportion of visits assigned to class 1, 2 and 3, based on the highest class membership probability, were 53.5% (n = 24,383), 29.8% (n = 13,596) and 16.7% (n = 7,603), respectively (Table 2). The degree of class separation (i.e., entropy) was 0.79, suggesting fairly high ability of the model to classify clusters [19]. The distributions of posteriori probabilities according to the assigned class memberships are provided in S1 Fig.

thumbnail
Table 2. Predicted probabilities (in percentages) of class membership and the outcome variables within each latent class.

https://doi.org/10.1371/journal.pone.0284056.t002

Sensitivity analyses stratifying by year to indicate potential longitudinal trends show very similar proportions of sexual behaviors over time with some small deviations in classes 2 and 3 (S2 Fig and S3 Table). Moreover, the class membership sizes (i.e., the number of visits assigned to each class) were quite stable in 2019 and 2020, but in 2021, the number of visits were notably higher in class 1 and lower in class 2.

STI diagnosis and sociodemographic variables associated with sexual behavior classes

An STI was diagnosed at 19.2% (n = 8,738) of all included visits. Per class, the percentage of visits at which an STI was diagnosed was 17.07% (n = 4,163) in class 1, 19.53% (n = 2,655) in class 2 and 25.25% (n = 1,920) in class 3. Table 3 shows that in multivariable analysis, with class 1 as the reference group, the visits classified in class 2 and class 3 were significantly more often with individuals who were diagnosed with an STI, older (≥36 vs. ≤35 years), MSMW (vs. MSM), and visiting an SHC in an urban (vs. non-urban) area. The odds ratios for being MSMW and visiting an SHC in an urban area were both notably high. Furthermore, the visits classified in classes 2 and 3, compared to those classified in class 1, were significantly less often with individuals who were from an STI/HIV endemic area. The visits in class 2 compared to class 1 had a significantly higher odds of being with individuals with a high (vs. low-middle) level of education. When comparing class 2 and 3 (S2 Table), the visits in class 3 (vs. class 2) were significantly more often with individuals diagnosed with an STI, aged ≥36 years and from an STI/HIV endemic area; while the odds of these visits were significantly lower in individuals who were MSMW (vs. MSM) and visiting an SHC in an urban (vs. non-urban) area.

thumbnail
Table 3. Association between class membership and various factors (LCA model with covariates) comparing classes 2 and 3 to class 1.

https://doi.org/10.1371/journal.pone.0284056.t003

In a sensitivity analysis, where the LCA model was repeated using only one randomly selected visit per individual, results were comparable to the original model (S4 Table); indicating that inclusion of multiple visits per participant did not likely bias results.

Discussion

This study has identified three latent classes of sexual behaviors among MSM, MSMW and TGP who are eligible for, but not using, PrEP. Most visits belonged to a subgroup of sexual behavior with a lower number of partners, and few reports of group sex, chemsex and sex work (i.e., class 1). Other visits belonged to subgroups of sexual behavior involving a higher number of partners and reported group sex (i.e., class 2), and to a lesser extent, of chemsex and sex work while also having numerous reports of high numbers of partners, group sex and sex work (i.e., class 3). The proportion of STI diagnosis increases with higher class number, suggesting that these subgroups represent one mean of assessing risk for acquiring STI, and thus also for HIV.

Individuals of all three classes were technically eligible for PrEP. However, the largest group, class 1, had a much lower proportion with STIs, likely related to different sexual behaviors compared to classes 2 and 3. Given that STI has been established as an important proxy for HIV infection [20, 21], individuals who exhibit the behaviors in class 1 at a given visit could be less prioritized for PrEP, despite being eligible, in favor of individuals who exhibit the behaviors in class 2 and 3. The identified risk groups based on sexual behavior profiles may therefore be more informative than simply PrEP eligibility in allocating PrEP to those at highest risk for HIV. In a situation of restricted subsidized PrEP provision, such as in the Netherlands, individuals presenting to the SHCs with sexual behaviors in line with class 3 should certainly be prioritized for PrEP.

When focusing on the components of classes 2 and 3, having more partners and engagement in group sex are predominant features of these latent classes and are both associated with increased HIV risk [4, 6]. One noteworthy difference between classes 2 and 3 is that at almost all visits in class 3, chemsex was reported. Even though chemsex is not in itself a transmission factor for HIV, it represents an environmental factor in which individuals are more at risk of exposure to HIV and condomless anal sex [22]. This finding is in line with a previous study using LCA in which classes including chemsex were associated with increased HIV and STI positivity among all MSM attending sexual health centers (SHC) in the Netherlands [23]. It should also be noted that specific combinations of chemsex drugs, such as erectile dysfunction drugs with nitrites and polydrug use, might be more closely linked to STI risk [24]. Moreover, we have limited data on whether chemsex includes injecting drug use due to substantial missing data. However, because it poses an increased risk for HIV acquisition, it is worth mentioning as another behavioral characteristic to consider for PrEP uptake [25]. Lastly, sex work was also an important feature for class 3; however, it was uncommon.

Individuals who presented at an SHC and belonged to class 2 or 3 were significantly more often 36 years old or older, MSM, not from an STI/HIV endemic area and visiting urban SHCs when compared to class 1. Notably, the demographic characteristics differentiating class 3 versus 2 were higher age (≥36 years old), lower proportion of MSMW, less likely to visit an urban SHC and more likely to be from an STI/HIV endemic area. As our data from the SHCs includes a noticeable proportion of individuals from an STI/HIV endemic area and non-urban SHC visitors, it can be assumed that although these individuals are not using PrEP, they are seeking care for their sexual health. Given their increased risk of STIs, as shown especially by the association with class 3, PrEP provision needs to be discussed and offered to these individuals from STI/HIV endemic countries and in non-urban areas [12, 26]; either within or outside of the NPP. Individuals from STI/HIV endemic regions were also more present in group 1 compared to other groups. This is despite their potential risk for HIV, as shown in research indicating that the majority of HIV acquisition among MSM migrants from Sub-Saharan Africa and Latin America/Caribbean are postmigration HIV infections, acquired in the European host country [27]. Nonetheless, those from STI/HIV endemic regions that were more often classified in group 1, represents individuals from heterogenous geographical locations. Given that we did not have access to specific information on the origin of these individuals or the lower risk of STI/HIV in the group with a larger proportion of migrants from endemic regions, this becomes difficult to interpret but important to consider for future research.

It is unknown whether the lack of PrEP use found among the eligible MSM reporting high proportions of sexual behaviors associated with STI/HIV (i.e., class 2 and 3) in our study may be attributed to not wanting PrEP, institutional rather than behavioral factors, such as being waitlisted for the NPP, or both. Interestingly, our analyses indicate that over the years, the proportion of visits increased in class 1, deceased in class 2, and remained stable in class 3. This shift in class membership is difficult to explain but could be due to changes in sexual behaviors due to COVID-19 restrictions [28, 29], the enrolment of individuals with high risk of HIV into the NPP over time, or other factors. The potential barriers for PrEP uptake may be alleviated by improving PrEP awareness, knowledge and providing alternative routes of affordable PrEP provision outside of the NPP such as telehealth. Further (qualitative) research is needed to gain a deeper understanding of why those in classes 2 and 3 are not using PrEP despite eligibility, patterns of PrEP uptake in relation to behavior and to what extent personal and institutional barriers play a role.

Our findings are based on a large sample size, using extensive surveillance data at a national level. Also, our data pertain to only visits at SHC. Even though this excludes those seeking sexual healthcare elsewhere, SHC visitors importantly reflect a population that is generally at increased risk of STI and HIV [3], thus for whom PrEP would more likely be indicated. Moreover, relatively few studies among general MSM populations report on those who do sex work, despite the importance of HIV prevention for this group. Lastly, because we have focused our study on behavior beyond the scope of the current PrEP eligibility criteria, we provide a more targeted way of allocating PrEP to those who need it most based on specific sexual behaviors and sociodemographic characteristics; especially when subsidized provision capacity is restricted.

Our study is limited in that we could not compare non-PrEP users to PrEP users. Previous research shows that these two groups did not differ sociodemographically, but PrEP users had more STIs, more sex partners, CAS and chemsex than non-PrEP users [30]. Because this comparison was not possible in our study, it is difficult to determine how the identified non-PrEP users differ from PrEP users and whether classes observed in this study are different among PrEP users. However, we do see that the STI prevalence is higher in classes 2 (20%) and 3 (25 = %) than among PrEP users in the NPP (17%) [3], indicating that those in classes 2 and 3 may have a similar or even higher need for PrEP than the average PrEP user in the NPP when considering the association between STI and HIV risk [20, 21].

Conclusions

In conclusion, this study has identified classes of sexual behavior that are associated with increased risk of STI and, as a strong correlate thereof, HIV. It is concerning that the individuals belonging to these classes who are visiting an SHC, are not using PrEP despite being eligible for it. We found that when many sexual partners, group sex, sex work or chemsex are reported, extra attention to encourage PrEP uptake is warranted. Prioritizing individuals with one or more of these characteristics may be attained through targeted information provision and counseling. In the case of restricted access and waitlists, an assessment for PrEP need based on the combination of these characteristics may be beneficial.

Supporting information

S1 Table. Comparison of fit statistics for latent class analysis with 1–6 classes.

Abbreviations: AIC = Akaike’s information criterion; BIC = Bayesian information criterion *Entropy could not be calculated.

https://doi.org/10.1371/journal.pone.0284056.s001

(DOCX)

S2 Table. Association between class membership and various factors (LCA model with covariates) comparing class 2 and 3.

Parameter estimates for the associations between STI diagnosis or sociodemographic variables and latent classes were directly obtained from a generalized structural equation model. Explanation of data: OR = odds ratio; aOR = adjusted odds ratio; 95% CI = 95% confidence interval. *aOR: all models were adjusted for the variables present in the table. **Any STI includes anal chlamydia, anal gonorrhea, hepatitis C virus, hepatitis B virus, and syphilis diagnosed at the visit. ***Originating from an STI/HIV endemic area is defined as being born in and having either one or both parents born in Suriname, Turkey, Netherlands Antilles, North Africa, Sub-Saharan Africa, Eastern Europe, Central and South America, or Asia. ****Region is defined as urban, referring to all “Randstad” provinces, or non-urban, referring to all other provinces.

https://doi.org/10.1371/journal.pone.0284056.s002

(DOCX)

S3 Table. Class membership size (number of visits) across three latent classes, stratified by year (2019, 2020 and 2021).

Explanation of data: Models were estimated using maximum likelihood, which was calculated by summing all conditional likelihoods of each latent class multiplied by the associated latent class probabilities. The posteriori probability of a visit i belonging to each class k, πik, was determined from this likelihood. Visits were then assigned a latent class k corresponding to the highest probability πik. Data are presented as percentages (n).

https://doi.org/10.1371/journal.pone.0284056.s003

(DOCX)

S4 Table. Association between class membership and various factors (LCA model with covariates) comparing class 2 and 3 to class 1; using one randomly selected visit per individual.

Parameter estimates for the associations between STI diagnosis or sociodemographic variables and latent classes were directly obtained from a generalized structural equation model. Explanation of data: OR = odds ratio; aOR = adjusted odds ratio; 95% CI = 95% confidence interval. *aOR: all models were adjusted for the variables present in the table. **Any STI includes anal chlamydia, anal gonorrhea, hepatitis C virus, hepatitis B virus, and syphilis diagnosed at the visit. ***Originating from an STI/HIV endemic area is defined as being born in and having either one or both parents born in Surinam, Turkey, Netherlands Antilles, North Africa, Sub-Saharan Africa, Eastern Europe, Central and South America, or Asia. ****Region is defined as urban, referring to all “Randstad” provinces, or non-urban, referring to all other provinces.

https://doi.org/10.1371/journal.pone.0284056.s004

(DOCX)

S1 Fig. The a posteriori probabilities of class membership per class.

Explanation of data: models were estimated using maximum likelihood, which was calculated by summing all conditional likelihoods of each latent class multiplied by the associated latent class probabilities. The posteriori probability of a visit i belonging to each class k, πik, was determined from this likelihood. Visits were then assigned a latent class k corresponding to the highest probability πik. The figures show the distribution of probabilities for belonging to a class given the assigned class membership. Each point represents an individual consultation visit. Figure A shows the probabilities for those assigned to class 1. Figure B shows the probabilities for those assigned to class 2. Figure C shows the probabilities for those assigned to class 3. For example, the dots in the figure indicate that only few visits had a lower probability of belonging to a given class, and the longer lines (subsequent dots), indicate that the majority of visits had (almost) 100% probability of belonging to a given class, and a very low probability of belonging to another latent class.

https://doi.org/10.1371/journal.pone.0284056.s005

(DOCX)

S2 Fig. Sexual behavior associated with increased STI risk across three latent classes, stratified by year (2019, 2020 and 2021).

Explanation of data: Bars represent the mean proportion of visits reporting each sexual behavior respectively for class 1, 2 and 3. All sexual behaviors refer to the six months prior to the visit. Number (= No.) of sexual partners refers to those with ≥6 sexual partners in the six months prior to the visit. Chemsex was defined as using cocaine, ketamine, mephedrone, gamma-hydroxybutyrate (GHB), gamma-butyrolactone (GBL), and/or crystal meth around or during sex. Bands at the top of each bar represent 95% confidence intervals, which were calculated using the delta method.

https://doi.org/10.1371/journal.pone.0284056.s006

(DOCX)

Acknowledgments

The authors gratefully acknowledge the nurses, physicians and supporting staff at the centres for sexual health, and the data managers and researchers at the RIVM for their contribution to the data collection.

References

  1. 1. Celum C, Baeten J. PrEP for HIV Prevention: Evidence, Global Scale-up, and Emerging Options. Cell Host & Microbe. 2020;27(4):502–6. pmid:32272075
  2. 2. Bierman W, Hoornenborg E, Nellen J, Janssens F, Burger D, Hermanussen R, et al. Nederlandse multidisciplinaire richtlijn Pre-expositie profylaxe (PrEP) ter preventie van hiv (update 2022) 2022. https://www.soaaids.nl/files/2022-07/20220711-PrEP-richtlijn-Nederland-versie-3-update-2022.pdf.
  3. 3. van Wees D, Visser M, van Aar F, de Coul EO, Staritsky L, Sarink D, et al. Sexually transmitted infections in the Netherlands in 2021. RIVM rapport 2022–0023. 2022.
  4. 4. Armstrong HL, Roth EA, Rich A, Lachowsky NJ, Cui Z, Sereda P, et al. Associations between sexual partner number and HIV risk behaviors: implications for HIV prevention efforts in a Treatment as Prevention (TasP) environment. AIDS care. 2018;30(10):1290–7.
  5. 5. Maxwell S, Shahmanesh M, Gafos M. Chemsex behaviours among men who have sex with men: A systematic review of the literature. International Journal of Drug Policy. 2019;63:74–89. pmid:30513473
  6. 6. Knox J, Boyd A, Matser A, Heijman T, Sandfort T, Davidovich U. Types of Group Sex and Their Association with Different Sexual Risk Behaviors Among HIV-Negative Men Who Have Sex with Men. Archives of Sexual Behavior. 2020;49(6):1995–2003. pmid:32500245
  7. 7. Goedel WC, Mimiaga MJ, King MRF, Safren SA, Mayer KH, Chan PA, et al. Potential Impact of Targeted HIV Pre-Exposure Prophylaxis Uptake Among Male Sex Workers. Scientific Reports. 2020;10(1):5650. pmid:32221469
  8. 8. Baral SD, Friedman MR, Geibel S, Rebe K, Bozhinov B, Diouf D, et al. Male sex workers: practices, contexts, and vulnerabilities for HIV acquisition and transmission. The Lancet. 2015;385(9964):260–73. pmid:25059939
  9. 9. Maxwell S, Gafos M, Moncrieff M, Shahmanesh M, Stirrup O. Pre-exposure prophylaxis use among men who have sex with men who have experienced problematic chemsex. Int J STD AIDS. 2020;31(5):474–80. pmid:32075538
  10. 10. Sundelson AE, Meunier É, Schrimshaw EW, Siegel K. Barriers to Pre-Exposure Prophylaxis Uptake Among Online Male Sex Workers in the US. AIDS and Behavior. 2022;26(5):1572–86. pmid:34705151
  11. 11. Hammoud MA, Vaccher S, Jin F, Bourne A, Maher L, Holt M, et al. HIV Pre-exposure Prophylaxis (PrEP) Uptake Among Gay and Bisexual Men in Australia and Factors Associated With the Nonuse of PrEP Among Eligible Men: Results From a Prospective Cohort Study. J Acquir Immune Defic Syndr. 2019;81(3):e73–e84. pmid:30973548
  12. 12. Annequin M, Villes V, Delabre RM, Alain T, Morel S, Michels D, et al. Are PrEP services in France reaching all those exposed to HIV who want to take PrEP? MSM respondents who are eligible but not using PrEP (EMIS 2017). AIDS Care. 2020;32(sup2):47–56. pmid:32189518
  13. 13. Keen P, Bavinton BR. Could disparities in PrEP uptake limit the public health benefit? Lancet Public Health. 2020;5(9):e467–e8. pmid:32888440
  14. 14. Staritsky L, Van Aar F, Visser M, Heijne J, Götz H, Nielen M, et al. Sexually transmitted infections in the Netherlands in 2019. 2020.
  15. 15. Stuart D. Chemsex: origins of the word, a history of the phenomenon and a respect to the culture. Drugs and Alcohol Today. 2019;19(1):3–10.
  16. 16. Centraal Bureau Statistiek (CBS). Definition Education level 2022 https://www.cbs.nl/en-gb/news/2018/20/well-being-not-distributed-equally/education-level].
  17. 17. Centraal Bureau Statistiek (CBS). Definition Migration background 2022 [https://www.cbs.nl/nl-nl/onze-diensten/methoden/begrippen/migratieachtergrond.
  18. 18. Skrondal A, Rabe-Hesketh S. Generalized latent variable modeling: Multilevel, longitudinal, and structural equation models: Chapman and Hall/CRC; 2004.
  19. 19. Celeux G, Soromenho G. An entropy criterion for assessing the number of clusters in a mixture model. Journal of Classification. 1996;13(2):195–212.
  20. 20. Hartwell TD, Pequegnat W, Moore JL, Parker CB, Strader LC, Green AM, et al. The utility of a composite biological endpoint in HIV/STI prevention trials. AIDS Behavior. 2013;17(9):2893–901. pmid:23748863
  21. 21. Cheung KT, Fairley CK, Read TR, Denham I, Fehler G, Bradshaw CS, et al. HIV incidence and predictors of incident HIV among men who have sex with men attending a sexual health clinic in Melbourne, Australia. PLoS One. 2016;11(5):e0156160. pmid:27219005
  22. 22. Basten MGJ, van Wees DA, Matser A, Boyd A, Rozhnova G, den Daas C, et al. Time for change: Transitions between HIV risk levels and determinants of behavior change in men who have sex with men. PLOS ONE. 2021;16(12):e0259913. pmid:34882698
  23. 23. Slurink IAL, van Benthem BHB, van Rooijen MS, Achterbergh RCA, van Aar F. Latent classes of sexual risk and corresponding STI and HIV positivity among MSM attending centres for sexual health in the Netherlands. Sexually Transmitted Infections. 2020;96(1):33. pmid:31221743
  24. 24. Achterbergh RCA, de Vries HJC, Boyd A, Davidovich U, Drückler S, Hoornenborg E, et al. Identification and characterization of latent classes based on drug use among men who have sex with men at risk of sexually transmitted infections in Amsterdam, the Netherlands. Addiction. 2020;115(1):121–33. pmid:31400174
  25. 25. Scheibein F, Wells J, Henriques S, Van Hout MC. “Slam Sex”—Sexualized Injecting Drug Use (“SIDU”) Amongst Men Who Have Sex with Men (MSM)—A Scoping Review. Journal of Homosexuality. 2021;68(14):2344–58. pmid:32875954
  26. 26. Bil JP, Zuure FR, Alvarez-del Arco D, Prins JM, Brinkman K, Leyten E, et al. Disparities in access to and use of HIV-related health services in the Netherlands by migrant status and sexual orientation: a cross-sectional study among people recently diagnosed with HIV infection. BMC Infectious Diseases. 2019;19(1):906. pmid:31664925
  27. 27. Alvarez-Del Arco D, Fakoya I, Thomadakis C, Pantazis N, Touloumi G, Gennotte A-F, et al. High levels of postmigration HIV acquisition within nine European countries. AIDS (London, England). 2017;31(14):1979–88. pmid:28857779
  28. 28. Jongen VW, Zimmermann HML, Boyd A, Hoornenborg E, van den Elshout MAM, Davidovich U, et al. Transient Changes in Preexposure Prophylaxis Use and Daily Sexual Behavior After the Implementation of COVID-19 Restrictions Among Men Who Have Sex With Men. J Acquir Immune Defic Syndr. 2021;87(5):1111–8. pmid:34229327
  29. 29. van Bilsen WPH, Zimmermann HML, Boyd A, Coyer L, van der Hoek L, Kootstra NA, et al. Sexual Behavior and Its Determinants During COVID-19 Restrictions Among Men Who Have Sex With Men in Amsterdam. J Acquir Immune Defic Syndr. 2021;86(3):288–96. pmid:33230027
  30. 30. Coyer L, Prins M, Davidovich U, van Bilsen WPH, Schim van der Loeff MF, Hoornenborg E, et al. Trends in Sexual Behavior and Sexually Transmitted Infections After Initiating Human Immunodeficiency Virus Pre-Exposure Prophylaxis in Men Who Have Sex with Men from Amsterdam, the Netherlands: A Longitudinal Exposure-Matched Study. AIDS Patient Care STDS. 2022;36(6):208–18. pmid:35687814