Among men who have sex with men diagnosed with early syphilis, the probability of notifying 1 or more sex partner did not differ by reported Internet-based app use to meet sex partners.
Correspondence: Victoria Mobley, North Carolina Division of Public Health, Department of Health and Human Services, 1902 Mail Service Center, Raleigh, NC 27699-1902. E-mail: [email protected].
Case reports of syphilis have been increasing in the United States (US) since 2013, with 2016 primary and secondary (ie, infectious) syphilis case rates at their highest since 1993.1 In 2016, North Carolina (NC) had the 8th highest infectious syphilis incidence rate in the US. As seen nationally, the majority of syphilis cases in NC are among men (86%), specifically men who have sex with men (MSM).1,2
Partner notification services (PNS) remains the backbone of syphilis prevention and control in NC. The goal of PNS is to interrupt disease transmission by notifying sex partners of confirmed cases, providing risk reduction counseling, and referral to testing and treatment for both cases and their sex partners. The NC public health law requires the Division of Public Health (NC DPH) to provide PNS for all individuals newly diagnosed with early syphilis (ES)-clinical or serologic evidence of having acquired the infection in the preceding 12 months. Public health officials responsible for performing PNS, disease intervention specialists (DIS), collect the names and locating information of all sex partners at risk of syphilis infections from the index case.3 Between 2013 and 2016, NC DIS located and interviewed 94% of all ES index cases.4
Despite locating and interviewing ES cases, the continued increase in NC’s syphilis morbidity suggests infected partners of index cases are not being identified or treated before disease transmission. The success of a PNS program depends on the index case’s willingness or ability to provide sufficient locating information for DIS to rapidly locate and treat their sex partners to prevent onward disease transmission. The increasing popularity of Internet-based applications (apps) to find sex partners, particularly among MSM5 can facilitate sex partner finding that does not require identification beyond a screen name, making the subsequent location of those partners difficult.6–9 Additionally, frequent changes in screen names or apps used to meet partners and governmental restrictions on DIS access to these apps may present barriers for successful PNS. Because data are needed to support programmatic and policy change, we performed a retrospective review of the surveillance data for all MSM ES cases reported in NC to assess the association between PNS outcomes and reported use of Internet-based apps to meet sex partners. Widely used PNS outcome indices were used to assure comparability of our observations to PNS assessments performed in other jurisdictions.
Surveillance data for syphilis are collected in the NC Electronic Disease Surveillance System. This system also contains DIS case investigative notes for primary, secondary and early latent syphilis index patients and their reported sex partners. We restricted our analysis to all ES MSM and their sex partners, reported to the NC DPH between January 1, 2013, and December 31, 2016. All patients not known to have late syphilis are assigned to DIS and prioritized for PNS. We classified male patients that self-reported being homosexual, bisexual, or reported a male sex partner during their infectious period as MSM.
During interviews with index patients, DIS elicit information about clinical manifestations of infection, sexual risk behaviors, the names and locating information of all sex partners during the patient's infectious period (primary, 3 months; secondary, 6 months; early latent, 12 months). Only sex partners reported by the index patient during their infectious window were included in our analysis.
As part of their investigation, DIS ask all index patients whether they met sex partners on an Internet-based app in the 12 months preceding their syphilis diagnosis. If an index patient answers yes to this question, DIS ask for the absolute number of both named and unnamed sex partners met on Internet-based apps during the patient’s infectious period (Internet sex partner). However, we do not know which specific partners were met online. We classified index patients who reported and did not report meeting 1 or more sex partner via an Internet-based app in the 12 months preceding their syphilis diagnosis as app users and non-app users, respectively. Patients who did not answer the Internet-based app use question were excluded from our analysis. Because the infectious period for primary and secondary syphilis is shorter than the period used to assign app user exposure (12 months), we estimated potential misclassification bias by calculating the proportion of patients who did not report 1 or more Internet sex partner during their infectious period.
Reported sex partners were stratified into 2 groups: named partners (index case provided enough locating information for the DIS to attempt location) and unnamed partners (index case did not provide enough locating information for the DIS to attempt location). Traditionally, a DIS needs at least one of the following pieces of locating information for the partner to initiate PNS: phone number, residential or employment address, or a social security number. Internet PNS (IPNS) refers to locating sex partners via Internet-based platforms. Locating information needed to initiate IPNS include an e-mail address, social media username, or app screen name. Sex partners met both online and offline could be eligible for traditional PNS and/or IPNS; the method of partner notification is dependent only on the types of locating information provided by the index patient to the DIS. Because IPNS for all syphilis investigations are performed by a single individual in NC, sex partners are assigned for IPNS when traditional PNS attempts have failed or when only IPNS identifiers are available.
From NC Electronic Disease Surveillance System, we extracted demographic, clinical, behavioral risk data, and syphilis treatment status information for ES index patients, notification of syphilis exposure, IPNS assignment, and syphilis testing and diagnosis status for named sex partners, and human immunodeficiency virus (HIV) status for both ES index patients and their named sex partners. We matched index patients and their named partners to the enhanced HIV/Acquired Immune Deficiency Syndrome Reporting System (eHARS) to determine their HIV status. Index patients were considered to have prevalent HIV if the HIV diagnosis date was before or within 60 days after the syphilis diagnosis date. Named sex partners were considered to have prevalent HIV infection if the HIV diagnosis date was before the PNS initiation date. Sex partners were considered newly diagnosed with HIV if the HIV diagnosis date was on or within the 30 days after the PNS initiation date.
Descriptive and bivariate analyses (Kruskal-Wallis for continuous variables and χ2 tests for categorical variables) were used to compare demographic (race/ethnicity and age), clinical (syphilis stage, syphilis treatment and HIV status) and behavioral (prior STI history, frequency of condom use, sexual risk practices, and success in getting patients to name partners) characteristics between app and non-app users. A P value less than 0.05 was considered statistically significant. We compared the following PNS outcomes between app and non-app users: partner notification index (number of named sex partners notified of their exposure per index patients interviewed), treatment index (number of named infected or exposed sex partners who received curative or empiric treatment for syphilis per index patients interviewed), and the case finding index (number of named sex partners newly diagnosed with syphilis or HIV infection per index patients interviewed).
Log binomial regression was used to estimate risk ratios (RR) and 95% confidence intervals (CI) to assess the association between app use and DIS notification of 1 or more sex partner of syphilis exposure and other index patient characteristics in univariate and multivariate models. A sensitivity analysis was performed to compare the probability of notifying all versus fewer than all named sex partners, by reported app use. Variables significantly associated with Internet sex partner finding in bivariate analysis were included in the multivariate models. Data were analyzed using STATA/IC 11.2 software (College Station, TX).
Between 2013 and 2016, 5560 individuals were reported with ES to the NC DPH. Most were male (87.1%, n = 4840), of which 73.9% (n = 3576) were MSM. Reported use of Internet-based apps to meet sex partners by MSM ES cases increased slightly throughout this analysis period (54.1% in 2013 to 59.3%; P = 0.13). DIS staff interviewed 3464 (96.7%) MSM ES patients, of which 3414 (98.6%) answered the Internet-based app use question; 1999 (58.6%, app users) answered “yes” and 1415 (41.4%, non-app users) answered “no.” The remaining 50 (1.4%) interviewed index patients were excluded from this analysis (Fig. 1). Excluded patients were more frequently older (median age, 34 years vs 29 years), reported no sex partners (22.0% vs 7.1%), had no PNS initiated sex partners (46.0% v 27.2%), and less frequently reported a new sex partner during their infectious period (39.1% vs 69.1%; all P's < 0.05) compared to those included.
Among app users, 87.4% of P&S index patients reported meeting 1 or more Internet sex partner during their infectious period. The most frequently reported apps used to meet sex partners were Adam4Adam (42.3%), Grindr (29.5%), Jack'd (20.7%), Facebook (9.5%), and Craigslist (5.5%). One third (33.5%) of app users and 24.9% of non-app users also reported meeting sex partners in bars or clubs. We were unable to determine where 75.1% of non-app users met their sex partners.
App using index patients were more frequently non-Hispanic, white (33.2% vs 27.3%; P = 0.003) and younger (median, 28 years vs 30 years; P = 0.0002) than non-app users (Table 1). App users were less likely than non-app users to be HIV coinfected (54.1% vs 58.2%; P = 0.02) to report a female sex partner (12.5% vs 15.5%; P = 0.01) or always using a condom during sexual intercourse (9.8% vs 13.6%; P = 0.002) and more likely to report a prior history of a STI (66.0% vs 62.2%; P = 0.008). The syphilis stage and the proportion of index cases that received treatment were similar for app and non-app users.
App users were more likely than non-app users to report 1 or more sex partner during their infectious period (94.9% vs 89.6%; P < 0.001) and less likely to have no sex partners initiated for PNS (27.7% vs 33.8%; P < 0.001). Among index patients who reported 1 or more sex partner, a similar proportion of app and non-app users reported only unnamed partners (23.6% vs 25.0%; P = 0.4). Among index patients with no initiated sex partners, app users were more likely than non-app users to report 1 or more new sex partner during their infectious period (80.3% vs 52.1%; P < 0.001) (Table 1). These observations did not change when stratified by HIV serostatus (data not shown).
A similar proportion of app and non-app users with 1 or more initiated sex partner had 1 or more sex partner notified of their exposure (81.4% vs 83.7%) and brought to treatment (41.1% vs 39.2%; all P's > 0.10). A smaller proportion of app users had 1 or more sex partner newly diagnosed with syphilis or HIV infection (10.0% vs 13.0%; P = 0.04). The median number of partners located (4 vs 2), treated for exposure or infection (4 vs 2) and newly diagnosed with syphilis or HIV infections (3 vs 2; all P's = 0.0001) were higher for app versus non-app users.
A total of 11,409 partners were reported by index patients; 8242 (72%) by app users and 3167 (28%) by non-app users. Nearly half of all partners reported in each a group were unnamed and thus not initiated for PNS (app users: n = 4278; 51.9%; non-app users: n = 1568; 49.5%), (Table 2). Internet sex partners comprised 41.5% (n = 3241) of partners reported by app users; nearly half of whom (46.9%) had enough locating information to initiate PNS or IPNS.
App users named more sex partners per index case interviewed than non-app users (1.98 vs 1.13; P = 0.0001), (Table 2). Similar proportions of partners named by app and non-app users had appropriate HIV follow-up: either a record of prevalent HIV infection (32.7% vs 34.2%, respectively) or a record of HIV testing as part of the DIS investigation (37.3% vs 39.6%; P's > 0.10). A lower proportion of partners named by app users received serologic testing for syphilis exposure (58.5% vs 63.2%; P = 0.001) or were newly diagnosed with syphilis (12.0% v 15.5%; P = 0.005). The proportion of partners tested for HIV that were newly diagnosed did not differ if named by an app user or non-app user (4.6% vs 3.3%; P = 0.31). Standard PNS outcomes were higher for app versus non-app users; partner notification index (1.57 vs 0.90), treatment index (0.80 vs 0.44), and case finding index (0.16 vs 0.12) (Table 1).
There were 613 (11.0%) named sex partners assigned for IPNS; 460 (11.6%) partners named by app users and 153 (9.6%) named by non-app users (P = 0.03) (Table 2). Disease intervention specialists assured the evaluation and treatment of approximately 35% of partners named by app and non-app users assigned for IPNS. The inclusion of IPNS activities increased the overall proportion of named partners that DIS successfully brought to evaluation and treatment from 49.3% to 53.4% for app users and from 48.3% to 51.7% for non-app users.
In total, 58.8% (n = 1176) app users and 55.4% (n = 784) non-app users named 1 or more partner notified by DIS of their exposure to syphilis. We observed no difference in the likelihood of the DIS being able to successfully notify 1 or more named partner of their exposure between app and non-app users in unadjusted (RR, 0.99; 95% CI, 0.88–1.12) or adjusted (RR, 0.99; 95% CI, 0.87–1.13) log binomial regression models (Table 3). In sensitivity analysis, the probability of notifying all versus fewer than all named partners did not differ between the two groups (data not shown).
In this analysis of MSM ES patients, the probability of notifying at least one named partner did not differ by reported use of Internet-based apps to meet sex partners. A similar proportion of both reported and named partners were located by DIS for app and non-app users, and nearly half of reported Internet sex partners had enough contact information to initiate PNS or IPNS. Due to the larger number of partners named by app users, those investigations identified and treated more partners than investigations of non-app users. Additionally, app and non-app using index patients reported similar proportions of unnamed sex partners, suggesting Internet-based app sex partner finding may not be the main or only factor limiting partner investigations for MSM index patients. Our analysis highlights the continued importance of PNS in locating and assuring the evaluation and treatment of named partners.
As observed previously,10–14 over half of MSM index patients reported using an Internet-based app to meet sex partners and Internet sex-seeking was correlated with larger numbers of reported sex partners per index patient. Although PNS was able to assure the evaluation and treatment of a similar percentage of named partners, app users still had almost 3 times as many partners that were either not located (both named or unnamed) or refused evaluation and treatment compared with non-app users. Additionally, as reported in other studies14–16 we observed a positive correlation between meeting sex partners on an Internet-based app and reported STI risk factors, such as a prior STD history, inconsistent condom use and engaging in sex during the syphilis infectious period. Elevated STI risk behaviors and numbers of reported partners among MSM who meet sex partners on Internet-based apps suggest websites used to find sex partners may provide a venue for increased syphilis transmission, and therefore may have contributed to the increasing syphilis epidemic in NC over the course of this analysis.
It is surprising that non-app users reported similar proportions of unnamed sex partners as app users and were more likely to report no sex partners during their infectious period. This may be partly explained by experience with DIS during this and previous interviews. Non-app users were more frequently HIV-positive, which has been previously associated with a lower number of named sex partners and PNS indices among MSM syphilis patients.10 Prior encounters with DIS at the time of HIV diagnosis may play a role in a patient's subsequent willingness to cooperate with PNS efforts. Qualitative research is underway in NC to identify, understand, and address the barriers MSM syphilis patients experience when reporting sex partners for PNS.
Incorporating IPNS with traditional PNS assured the evaluation and treatment of a similar proportion of partners named by app and non-app users, improving overall PNS outcomes for both groups. Our findings demonstrate the expanded benefit that IPNS could provide in reaching populations who routinely use Internet-based platforms to communicate. Unfortunately, like many jurisdictions,17 and due to legal concerns in NC regarding the use of government-owned electronic devices to access websites and apps used by patients to meet sex partners, state-employed DIS staff cannot directly perform IPNS. Restricting IPNS activities to one contracted employee has also limited our ability to assess the value IPNS could have if it were an integrated part of all DIS investigations. In addition to enhancing disease intervention efforts, authorizing DIS to access these Internet apps and websites for IPNS purposes would ensure DIS maintain cultural competency in the rapidly changing cyber landscape where MSM socialize and meet sex partners.
The routine capture of the number and classification of all sex partners reported during DIS interviews allowed us to assess the contribution of unnamed sex partners to NC’s syphilis epidemic. All index patient risk behavior and partner information obtained during DIS interviews were self-reported and therefore subject to recall or response bias. Limitations in data collection prohibited the identification of which sex partners were met online, therefore we cannot comment on differences in PNS outcomes by where partners were met. Some misclassification of app use exposure was possible among primary and secondary syphilis patients because the infectious period was shorter than the 12-month period used to assess Internet sex partner finding. However, we believe this misclassification to be minimal because most of these patients reported at least one Internet sex partner during their infectious period.
Given the high proportion of unnamed partners reported by index ES MSM patients, regardless of app usage, further study is necessary to understand how to provide PNS that is supportive of MSM culture and effective for notifying and treating partners. Though Internet-based app use to find sex partners was associated with more overall sex partners, reported STI risk behaviors and infected partners, we observed no difference in the proportion of named partners notified of exposure. Our analysis highlights the continued usefulness of PNS for syphilis intervention and the potential benefit that routinizing IPNS into DIS investigations could have on syphilis partner notification outcomes for MSM regardless of app use to meet sex partners.
1. Centers for Disease Control and Prevention. 2016. Sexually Transmitted Diseases Surveillance. Available at: https://www.cdc.gov/std/stats16/Syphilis.htm
. Accessed September 13, 2017.
2. Health, North Carolina Department of Health and Human Services/ Division of Public. N.C. Communicable Disease Reports. 2016 North Carolina HIV/STD/Hepatitis Surveillance Report. Available at: http://epi.publichealth.nc.gov/cd/stds/figures.html
. Accessed September 13, 2017.
3. North Carolina Administrative Code. 10A NCAC 41A .0204 Control Measures-Sexually Transmitted Diseases. Available at: reports.oah.state.nc.us/ncac/title%2010a%20…/10a%20ncac%2041a%20.0204.pdf
. Accessed September 13, 2017.
4. Field Services Unit Performance Report -Syphilis Index Cases Assigned/Interviewed 2013–2016. North Carolina electronic disease surveillance system (database online). Report run September 1, 2017.
5. Grov C, Breslow AS, Newcomb ME, et al. Gay and bisexual men's use of the internet: Research from the 1990s through 2013. J Sex Res 2014; 51:390–409.
6. Li Q, Liu Y, Zhou Z, et al. Online sex-seeking behaviors among men who have sex with men: Implications for investigation and intervention. AIDS Behav 2012; 16:1690–1698.
7. Wang AL, Peng RR, Tucker JD, et al. Optimizing partner notification programs for men who have sex with men: Factorial survey results from South China. PLoS One 2016; 11:e0157749.
8. Tan WS, Chen M, Ivan M, et al. Partner notification outcomes for men who have sex with men diagnosed with syphilis referred to partner notification officers, Melbourne, Australia. Sex Transm Dis 2016; 43:685–689.
9. van Aar F, van Weert Y, Spijker R, et al. Partner notification among men who have sex with men and heterosexuals with STI/HIV: Different outcomes and challenges. Int J STD AIDS 2015; 26:565–573.
10. DeSilva M, Hedberg K, Robinson B, et al. A case-control study evaluating the role of Internet meet-up sites and mobile telephone applications in influencing a syphilis outbreak: Multnomah County, Oregon, USA 2014. Sex Transm Infect 2016; 92:353–358.
11. Rowlinson E, Goings S, Minnerly S, et al. Differences in partner services outcomes for men WHO have sex with men diagnosed with primary and secondary syphilis by HIV serostatus. Sex Transm Dis 2018; 45:152–157.
12. Hochberg CH, Berringer K, Schneider JA. Next-generation methods for HIV partner services: A systematic review. Sex Transm Dis 2015; 42:533–539.
13. Klausner JD, Wolf W, Fischer-Ponce L, et al. Tracing a syphilis outbreak through cyberspace. JAMA 2000; 284:447–449.
14. Liau A, Millett G, Marks G. Meta-analytic examination of online sex-seeking and sexual risk behavior among men who have sex with men. Sex Transm Dis 2006; 33:576–584.
15. Taylor M, Aynalem G, Bemis C, et al. Correlates of internet use to meet sex partners among men who have sex with men diagnosed with early syphilis in Los Angeles County. Sex Transm Dis 2004; 31:552–556.
16. Zou H, Fan S. Characteristics of men who have sex with men who use Smartphone Geosocial networking applications and implications for HIV interventions: A systematic review and meta-analysis. Arch Sex Behav 2016; 46:885–894.
17. Golden MR, Katz DA, Dombrowski JC. Modernizing field services for human immunodeficiency virus and sexually transmitted infections in the United States. Sex Transm Dis 2017; 44:599–607.