Search

Background As part of an ongoing sexual network study, we assessed the impact of COVID-19 on sexual behavior of men who have sex with men (MSM). Methods The Columbus, Ohio arm of the multi-site Network Epidemiology of Syphilis Transmission (NEST) study is following 241 MSM over two years. Participants attend quarterly visits with behavioral surveys and HIV/STI testing. In April 2020, we implemented an additional survey to measure the impact of COVID-19 on sexual behavior. Our analysis compared participants’ reported sexual behavior during two periods: April-July 2020 (T1) and August-December 2020 (T2). Results Of the 200 participants who completed at least one COVID-19 survey in either time period, 10 (5%) self-reported a previous COVID-19 diagnosis. We compared behaviors reported by 174 respondents who submitted surveys in both time periods. During T1, about half (46%) said they felt more anxious about sex since COVID-19; anxiety was similar during T2 (43%). Many reported engaging in online sexual activities, though we saw no meaningful differences by time period for starting or increasing chatting on hookup apps (T1=30%, T2=30%), sexting (T1=30%, T2=25%), or video chat sex (T1=16%, T2=14%). Watching pornography was reported more often in T1 than T2 (53% vs. 42%). Nearly a third of respondents (30%) in T1 reported that they were not having sex; this decreased to 17% in T2. Compared to pre-pandemic behavior, in T1 a majority of participants (74%) said they were less likely to have sex with a new partner, compared to 61% during T2. Conclusion During COVID-19, nearly half of MSM reported feeling more anxious about sex, and many turned to online activities in lieu of in-person sex. However, when comparing the early months of the pandemic to later months of 2020, more participants are resuming in-person sex, and willingness to have sex with new partners also rose.

Background Estimation of epidemic potential of a sexually transmitted infection (STI) is difficult due to difficulty in measuring and quantifying the sexual network and implications for infection transmission. We demonstrate an approach for predicting the epidemic potential of an STI using data on another STI for men who have sex with men. Methods An individual-based mathematical model was constructed to describe sex partnering and STI concurrent transmission, namely HIV, herpes simplex virus type 2 (HSV-2), gonorrhea, chlamydia, and syphilis. The model was parameterized with representative biological and behavioral data. 500 heterogeneous sexual networks were simulated, on each of which STI transmission was also simulated. Correlations were assessed on model simulations (STI prevalences). Regressions were conducted to evaluate the predictability of HIV prevalence from each of the other STI prevalences. Results Across these simulations, Spearman’s rank correlation coefficient was 0.46 (95% CI: 0.37- 0.55) between HIV and HSV-2, 0.90 (95% CI: 0.88–0.91) between HIV and gonorrhea, 0.82 (95% CI: 0.78–0.86) between HIV and chlamydia, 0.82 (95% CI: 0.78–0.84) between HIV and syphilis, 0.31 (95% CI: 0.21–0.40) between HSV-2 and gonorrhea, 0.82 (95% CI: 0.78–0.86) between HSV-2 and chlamydia, 0.15 (95% CI: 0.05–0.25) between HSV-2 and syphilis, 0.70 (95% CI: 0.65–0.75) between gonorrhea and chlamydia, 0.93 (95% CI: 0.92–0.95) between gonorrhea and syphilis, and 0.56 (95% CI: 0.49–0.61) between chlamydia and syphilis. The adjusted R-squared for predicting HIV prevalence using each individual STI prevalence was 0.40 for HSV-2, 0.77 for gonorrhea, 0.71 for chlamydia, and 0.57 for syphilis. The adjusted R-squared for predicting HIV prevalence in a model that includes all other STI prevalences was 0.92. Conclusion STI prevalence is a proxy biomarker of HIV prevalence across heterogeneous sexual networks, explaining a considerable fraction of HIV prevalence variation. However, the strength of the association between each pair of STIs varies across STIs. Disclosure No significant relationships.

Background Network position contributes to STI risk because of broad sexual network connectivity. Relationship types (i.e., exclusive dating, hook ups) also influence STI risk. Few studies examine how sexual network position and relationship types jointly influence individuals’ STI risk reduction strategies. Methods We analyzed data from 2011–2017 National Survey of Family Growth for 5,948 and 5,433 unmarried, non-cohabiting sexually active women and men aged 15–44. To describe sexual network position, we created four dichotomous variables that included both past-year number of opposite-sex sex partners (one, multiple) and perceived partner non-monogamy (PPNM, yes/no). Relationship type was defined as a steady or casual sex partner at last sex. Prevalence ratios were used to assess the relationship between network position and two past-year outcomes: condom use at last sex and STI testing. Results were stratified by relationship type. Results 54% and 47% of women and men aged 15–44 had one partner with no PPNM. 6% and 4% had one partner with PPNM, 22% and 29% had multiple partners with no PPNM and 17% and 20% had multiple partners with PPNM. Men with multiple partners and PPNM had the lowest prevalence of condom use of all four groups (37.7%, CI: 33.5, 44.1 compared with 52–56%). Moreover, women and men in this group with a steady sex partner had higher prevalence of past-year STI testing if they reported PPNM than if they did not (women: aPR=1.31; CI: 1.11, 1.55, men: aPR=1.47; CI: 1.19, 1.81). This same relationship was not seen for women and men whose last sex was with a casual partner. Conclusion Individuals’ STI risk reduction strategies depend on sexual network position and relationship type. Those with PPNM and a steady sex partner may seek STI testing more frequently. Proxy sexual network measures from national surveys may help target STI prevention and testing interventions. Disclosure No significant relationships.

Background Use of the internet to meet sex partners has increased over time among MSM, and an outbreak of syphilis has been linked to seeking sex partners through an online chatroom. Other studies have shown associations between methamphetamine use(MU), sexual risk behaviors and STIs including syphilis. The objective was to determine the association between syphilis positivity, MU, sexual risk behaviors and sex partner meeting spaces and separately, MU, sexual risk behaviors and sex partner meeting spaces among Black MSM(BMSM) in one mid-Atlantic U.S. city without a known history of these associations. Methods Data came from an ongoing longitudinal cohort study, the Understanding Sexual Health in Networks Study (USHINE) including MSM 18–45 years of age in Baltimore City. Information on MU and sex partners and meeting spaces in the past 3 months was obtained via an egocentric sexual network survey. Summary statistics, chi-squared tests and t-tests were used for hypothesis testing. Results 147 MSM were enrolled and 75.5% (111) were BMSM. The mean age of participants was 29.7 years (SD 5.60); 33.3%(34) were syphilis positive. Syphilis positivity was associated with MU (57.1% vs 29.6%, p-value=0.042) and meeting sex partners on Jack’d (42.3% vs 23.4%, p-value=0.046) and Facebook (52% vs 27%, p-value=0.022). MU (vs no use) was associated with anonymous sex (69.2% vs 19.3%, p-value Conclusion Among BMSM, syphilis infection was associated with MU and specific sex partner meeting spaces, and MU was associated with sexual risk behaviors and specific sex partner meeting spaces. The relationships have not been identified before in this setting and suggest specific meeting spaces may be important access points for syphilis and drug use prevention. Disclosure No significant relationships.

Background A combination of effective and efficient approaches are necessary for scaling up of HIV case-identification, particularly in resource-limited settings. To increase access and coverage, communities need to be linked to facilities via community-based interventions that seek to promote health seeking behavior. This study is a Comparative Analysis of Facility Optimization and Community Based HIV Intervention and compares the effect of community HIV testing services(HTS) to HTS optimization at the facility-level. Methods This is a pre-and post-intervention study conducted in Eleme, one of the priority LGA supported by the USAID funded Strengthening Integrated Delivery of HIV/AIDS Services (SIDHAS) in Rivers State. The pre-intervention phase (PIP1) covers the period November 2015 – October 2016 while the post intervention phase is from November 2016 – Oct 2017. PIP1 involved community entry/mobilization, HIV screening in general population, referrals and linkage to care and treatment services from the community to the facility, while the PIP2 focused on optimization of HIV testing services within the facilities through multipoint/Provider Initiated Testing and Counselling[PITC], targeted testing in the communities, Sexual Network and Genealogy Testing and referrals by escort to Service Delivery Points. We reviewed HTS and ART commencement data to compare differences in positivity yield and linkage between both phases. Results The PIP1 had 107,813 individuals counselled, tested and received result, 1,406 tested HIV Positive and 964 linked to ART while the PIP2 had 24,078 individuals tested, 614 HIV positive and 610 linked to ART. Findings show increase in positivity yield from 1% to 3% and linkage from 87% to 99% in PIP1 and PIP2 respectively. Conclusion Although community outreaches create awareness, a targeted approach to HTS including sexual network/genealogy testing may be a more efficient approach. In addition, PITC in health facilities yields a higher positivity and linkage rates, maximizes use of testing resources by focusing on higher risk populations. Disclosure No significant relationships.

Background We describe a sexual network consisting of two males and five females who were tested for gonorrhoea at genital and oropharyngeal sites in early 2018, with the aim of exploring whether tongue kissing transmits gonorrhoea. Methods Participants independently provided extremely detailed interview accounts of sexual activity with one another, to permit inter-participant verification. Testing for N. gonorrhoeae was by nucleic acid amplification testing (NAAT). Whole genome sequencing (WGS) was performed on available culture-positive samples to assess genomic relatedness between samples. Results Sexual contact included tongue kissing for all participants, and many participants participated in oro-genital and genito-genital intercourse with one another. All sexual contact occurred during a 1-week period in Victoria, Australia. Two participants had samples available for WGS and these were highly related genomically, and these participants were separated in this network by two other participants. All seven participants had gonorrhoea, six participants had oropharyngeal gonorrhoea in the absence of genital gonorrhoea, and the other participant had only vaginal gonorrhoea. No men acquired urethral gonorrhoea. Conclusion The fact that six of seven participants had oropharyngeal gonorrhoea in the absence of urogenital gonorrhoea supports the notion that tongue kissing is a common mode of gonorrhoea transmission. No men acquired urethral gonorrhoea from oro-penile sex, suggesting that transmission from oropharynx to urethra may be less likely than by tongue kissing. Our WGS results, when viewed in conjunction with the extremely-detailed sexual behaviour data, support the notion that gonorrhoea diagnoses in this network were likely the result of within-network transmission. Our findings highlight the need for more research to define the transmission routes for gonorrhoea in heterosexuals, to inform screening policies. Disclosure No significant relationships.

Background Alongside traditional epidemiologic tools, network analyses and molecular epidemiology offer deeper insights into the structure of STBBI epidemics. In the context of a 2014 gonorrhea (NG) outbreak, this study sought to compare molecular networks to case-contact networks constructed from public health investigations. Methods Data were from enhanced public health investigations of NG in Winnipeg, Canada. NG-MAST was used to determine the molecular subtypes of NG. Subtypes were described by socio-demographic/clinical characteristics. Multivariable logistic regression models were used to assess the association of socio-demographic/clinical characteristics, and having the most frequently reported subtype. Networks constructed from case-contact investigations were visualized; components were characterized with univariate network statistics, including degree centralization. Conditional uniform graph (CUG) tests assessed observed degree centralization. Results In total, 126 NG cases were genotyped, with 41 subtypes found. Five subtypes accounted for 51% of all subtypes, with ST-3672 (n=22) predominant. At the bivariate level, infection with ST-3672 was associated with younger age (62% of those infected were 15–19 years old, p=0.002), and chlamydia co-infection (67% vs 37%, p=0.012). In multivariable analysis, age group remained significant, while an interaction between inner-core residency and chlamydia co-infection was detected. Case-contact networks were highly-fragmented, consisting mainly of dyads and triads. Of 85 components, the largest component included 6 nodes, while 61% were dyads. CUG testing indicated in-degree centralization was lower than expected (p Conclusion Molecular data revealed connections that were not apparent from case-contact investigations alone, leading to more cases potentially linked together, and over a wider geographic area. A handful of subtypes were responsible for the majority of infections. Early identification of dominant subtypes may potentially curtail transmission of NG. Disclosure No significant relationships.

Background HIV and other sexually transmitted infections (STIs) do not operate in isolation, particularly as people with risk-taking sexual behaviour may be co-infected. In this complex landscape, policy makers are limited by resource constraints while trying to find optimal coverage solutions. Disease modelling could help in this context. We aim to develop a user-friendly modelling software examining several STIs and HIV simultaneously, as we are unaware of any multi-STI decision support tools currently available. Methods We developed STI modelling software using the programming language Java, consisting of several models and a graphical user interface (UI). The models were drafted based on literature reviews and subsequently refined by experts, e.g. STI clinicians and policy makers. All models were internally and externally validated. The UI was developed with UI development experts and policy makers. Results Separate disease models, which describe the progression of chlamydia, gonorrhoea, HIV, syphilis, and their sequelae are included in the software. Sexual network models are used to describe the formation and dissolution of partnerships and thereby the occurrence of sexual contacts. Four different network models are included in the software. The clinical pathway models describe interventions, for example screening or STI treatment and reflect the current UK setting. All the models are interacting, individual-based discrete event simulations and have been validated using sensitivity analyses and publicly available data sources. The UI has been validated by policy makers. Conclusion With this modelling software policy makers can compare both existing and hypothetical intervention options in regards to their costs and consequences. All the parameters, formulas, model structures, and clinical pathways are editable. The software is flexible and usable in different settings and contexts. It can be and updated if needed, e.g. if medical knowledge changes. By adapting parameters which describe treatment pathways the software could be used in non-UK settings. Disclosure No significant relationships.

Background U.S. chlamydia and gonorrhea rates have increased for four consecutive years, reaching record highs in 2017. Similar trends were reported for the U.S. Army, with Army rates being elevated relative to the general population, due in part to demographic differences. A comparison of standardized rates was needed to put these differences in perspective. Methods Incidence rates of chlamydia and gonorrhea reported during 2013–2017 among Army active members and U.S. citizens age 15–64 were standardized using the 2015 Army age and sex distribution. The CDC’s National Electronic Disease Surveillance System and the Army’s Disease Reporting System Internet were used for the analysis. Results Crude and adjusted chlamydia rates (per 100,000) were over 2-fold higher among Army members (adjusted 2017 rates: 2,160 and 1,005 in the Army and U.S, respectively). Army chlamydia rates were elevated for all age and sex strata. The Army’s crude gonorrhea rates (per 100,000) were elevated (annual rate range: 275–373 versus 156 to 266 in the U.S.); however, adjusted U.S. rates surpassed Army rates (2017 rates: 438 in the U.S. versus 360 in the Army). Elevations for chlamydia and gonorrhea were observed in Army women under 25 relative to U.S. women 15–24 (2017 crude rates: 11,132 versus 3,635, respectively, for chlamydia and 1,117 vs 623, respectively, for gonorrhea). Crude gonorrhea rates were higher in U.S. men 25–44 relative to Army peers (2017 rates: 542 vs 269 for men 25–34, and 238 vs 106 for men 35–44, respectively). Conclusion The Army’s incidence of chlamydia was elevated relative to the general population even when demographic differences were taken into account. This may reflect higher individual or sexual network risks or better access to care. The Army’s lower adjusted gonorrhea rates may reflect differences in high-risk sub groups such as MSM, differing sexual network risks, or unmeasured confounders. Disclosure No significant relationships.

Background Some studies have shown an association between HIV pre-exposure prophylaxis (PrEP) use, STD acquisition and increased sexual risk behaviors (e.g. condomless sex). The objective was to determine the association between PrEP use, STD infection (i.e. syphilis, gonorrhea, chlamydia) and sexual risk behaviors (i.e. condomless sex, sex partner concurrency) in one mid-Atlantic city with an established epidemic of HIV. Methods Data came from the Understanding Sexual Health in Networks Study (USHINE), an ongoing longitudinal cohort study of MSM between the ages of 18–45. Participants completed an egocentric sexual network survey. Summary statistics, chi-squared tests, and t-tests were used for hypothesis testing. Results 173 men completed the sexual network survey. 52 (30.1%) men were HIV uninfected and 52 (30.1%) men reported PrEP use. PrEP users (vs non-users) were less likely to be positive for syphilis (10.6% vs 26.8%, p=0.03), but no more likely to be positive for gonorrhea (15.4% vs 8.0%, p=0.25) or chlamydia (17.3% vs 16.0%, p= 0.86). PrEP users (vs non-users) were more likely to report condomless sex (88.5% vs 69.2%, p =0.02) and sex partner concurrency (57.7% vs 32.7%, p=0.01). Conclusion PrEP users were less likely to be infected with syphilis and no more likely than non-PrEP users to be gonorrhea and chlamydia infected. PrEP users were more likely to report sexual risk behaviors including condomless sex and sex partner concurrency compared to non-PrEP users. The increased sexual risk behaviors among PrEP users is concerning and emphasizes the importance of prevention messaging for PrEP users. Disclosure No significant relationships.

Background Sexually transmitted infections (STIs) are a consistent medical concern among military service members, who may experience select rates well above their civilian counterparts. Differences in individual risk behaviors as well as network risk factors are different in military populations which may explain the increased rates of select STIs. We performed a pilot study to validate a STI risk assessment survey tool for future use in the military population. Methods Naval Medical Center Portsmouth was used to pilot a survey for a larger, egocentric social, and sexual network study in military populations. Study participants (n=50) completed an anonymized computer-assisted survey focusing on STI risk factors, including a sexual partner inventory. We evaluated individual demographics, behavioral risk factors, and sexual partnership formation to include mixing patterns and concurrency. Recent history of STI was self-reported in the survey and validated with the participant’s medical record. Fischer’s exact test was used if applicable. For partners, data was weighted and Wald Chi-square test was used to assess association between variables. Results Completion of the survey and sexual inventory by respondents was common. The majority were active duty, enlisted, and male. The survey suggested evidence of concurrency and disassortive mixing by age, race, and active duty status. Respondent (p=0.0089) and sexual partner (p=0.0401) alcohol consumption before sex was common and associated with history of STI. Condom use was inconsistent and less frequent with main/steady partners when compared to casual and anonymous partners. Conclusion The high completion rate demonstrated in this pilot study support that a military population will complete a detailed STI risk index including sexual partner inventory. Interim data analysis suggests that common individual risk factors such as network/partnership risk factors can be analyzed using an egocentric survey design. These results will have future implications for targeted prevention programming in the U.S. Military. Disclosure No significant relationships.

Background Swingers, i.e. heterosexual couples who, as a couple, have sex with others within the swingers’ subculture, are a subpopulation at risk for Sexually Transmitted Infections(STI). Human papillomavirus (HPV) has proven to be highly prevalent in female swingers (unpublished data). The aim of this study is to assess how the different HPV genotypes are distributed within the sexual networks of swingers and whether certain HPV genotypes cluster. Methods We build a sexual network of swingers based on the network data given by 115 swingers who participated in our prospective cohort network study on swingers between April 2011 and March 2012. The vaginal samples of all female swingers in the network and known in our STI clinic were tested for HPV. HPV genotypes were identified by the highly sensitive SPF10-LiPA25 method detecting 25 different genital HPV genotypes. The multilevel transmission network consisted of the one-mode sexual network of the swingers and the two-mode network of swingers and the HPV genotypes. Distribution and clustering of HPV genotypes in the sexual network was assessed using multilevel network analysis. Results The sexual network contained 1231 individual swingers of whom 556 female swingers. The sexual network consists of 38 separate components. Of 103 female swingers known in our STI clinic and tested for HPV, 96 (93%) were positive for at least one of the 25 HPV genotypes. All 25 HPV genotypes were present, but not equally distributed throughout the network. Some genotypes clustered together in specific components of the sexual network. Conclusion All assessed genotypes of HPV were prevalent in the sexual networks of swingers. Despite the incomplete data in our study, multilevel network analysis turns out to be useful for studying transmission networks. Clustering of highly prevalent HPV could be used as a marker for sexual behaviour, potential STI risk and transmission in swingers. Disclosure No significant relationships.

Background There are extreme racial disparities in HIV infection among MSM in the U.S. These disparities may in part be explained by racial differences in the transmission potential (i.e. mixing between infected and uninfected individuals) and structure (i.e. density) of sexual networks. The objective was to determine whether the association between HIV serodiscordance and sex partner concurrency differed by race among MSM sex partner dyads. Methods Data came from the Understanding Sexual Health in Networks Study, an ongoing longitudinal cohort among MSM ages 18–45. Participants completed an egocentric sexual network survey with questions about 3 most recent sex partners in the past 3 months. An HIV serodiscordant partnership was defined as a dyad with a positive status index and a negative or unknown status partner or vice versa. Summary statistics, chi-squared tests, and logistic regressions adjusted for the nested structure of the data were used for hypothesis testing. Results 163 MSM reported on 354 (median: 3, range: 0–3) sex partnerships. MSM were 63.2% Black (BMSM), on average 29.4 (SD 5.96) years old, and 33.6% reported condom use at last sex. There were no differences in age or condom use by race. Among partnerships, index BMSM (vs. non-BMSM) were more likely to report serodiscordant partnerships (48.8% vs. 14.4%, p-value Conclusion Among BMSM dyads, we found evidence of the necessary and sufficient causes for HIV transmission including mixing between infected and uninfected individuals (i.e. serodiscordance) combined with dense sexual network structures (i.e. sex partner concurrency) and we did not find this evidence among non-BMSM dyads. These factors may help explain persistent racial disparities in HIV. Disclosure No significant relationships.

Background In preparation for a prospective network study of men who have sex with men (MSM), we conducted a mixed-methods pilot study to examine feasibility and acceptability of several study components, including recruitment, data collection, and compensation. Methods We conducted in-depth interviews (IDIs) with eight providers serving MSM and five focus group discussions (FGDs) with 34 MSM from four target MSM populations: young Black men, HIV-positive men, HIV-negative men on pre-exposure prophylaxis (PrEP), and men not engaged in medical care. We also conducted a 4-week pilot of a smartphone app for ecological momentary assessment (EMA) with 20 MSM (ongoing). EMA surveys were employed to capture daily data on anal and oral sex, consumption of alcohol and drugs, use of hook-up and social networking apps, and other behaviors. We used a brief exit survey to assess EMA app acceptability. Results A major theme identified during IDIs and FGDs was the importance of developing trust and maintaining confidentiality during proposed respondent driven sampling (RDS) recruitment activities. A second FGD theme emphasized the importance of compensating participants appropriately for RDS and other study activities. All EMA participants reported being “completely comfortable” reporting their sexual behavior through the app. Most (67%) preferred the app to in-person interviews. Several participants identified technical issues with the app, including not receiving push notifications and spontaneous app closure. Conclusion IDIs and FGDs confirmed that developing trust and protecting participant confidentiality are critical for successful RDS recruitment. FGDs showed that MSM value their contributions to research and desire commensurate compensation. The EMA app was acceptable, despite technological challenges. It is feasible to use EMA to capture sexual behavior in this population. This mixed-methods pilot allowed for adjustments to the planned network study, including changes in compensation type and amount, troubleshooting technical issues, and modifications to the EMA survey. Disclosure No significant relationships.