Koumans, Emilia H. MD, MPH*; Farley, Thomas A. MD†; Gibson, James J. MD‡; Langley, Carol MD, MPH,§; Ross, Michael W. PhD, MPH,‖‖; McFarlane, Mary PhD*; Braxton, Jimmy*; St Louis, Michael E. MD*
IN 1998, RATES OF SYPHILIS in the United States declined to the lowest rate since data were first collected. 1 Nevertheless, syphilis transmission persists in inner cities and in certain areas of the southern United States. 2 Persistent syphilis in these areas can serve as the reservoir for the reemergence of epidemic syphilis, and highlights the inability of traditional sexually transmitted disease (STD) control efforts to eliminate this disease. 3 Because syphilis facilitates HIV infection 4 and these two infections overlap extensively at the ecologic 5 and individual levels, 6 eliminating syphilis is not only important to prevent the sequelae of syphilis but also to control HIV infection.
Epidemics of syphilis in the last 20 years have been associated with crack cocaine use and exchanging money and drugs for sex. 7–9 These behaviors have been postulated to facilitate syphilis transmission and hinder local control efforts because of the large number of partners and anonymous sexual contacts which are difficult to trace and, thus, allow for increased subsequent transmission. While determinants of epidemic syphilis have been carefully studied and mathematically modeled, 10,11 determinants of low rates but persistent syphilis transmission are less well understood. 12,13 What are the factors associated with persistent transmission of syphilis? Are the same factors responsible both for recent epidemics of syphilis and for persistent syphilis transmission? How is transmission of syphilis maintained in areas with persistent syphilis?
We studied syphilis in four areas with persistent transmission (Figure 1) to identify characteristics of persons who might have a role in the persistence of syphilis. Mathematical models of hypothetical populations have identified some behaviors that maintain or increase STD or HIV transmission. 14–17 Sexual network descriptions have also illustrated how STD and HIV 18–20 transmission is maintained between sexually connected persons. 21 Having concurrent or overlapping partners—defined as more than one partner during a specified period of time with the presumption that sexual partners are alternated—is a common factor considered in mathematical models and is viewed as potentially important in maintaining transmission in sexual networks. Many people are connected through concurrent sexual partnerships, and since sexual partners may also have concurrent sexual partners, a sexual partnership network is formed. However, the role of concurrent partners in the transmission of syphilis has not yet been examined or validated with empiric data from infected individuals, nor has this behavior been used to direct STD control activities. 22
Beginning in 1995, the research project Innovations in Syphilis Prevention (ISP) has addressed syphilis in areas in the southern United States that historically have had persistent syphilis transmission. We identified individuals who may contribute to sustained transmission of syphilis, and characterize demographics and behavioral features associated with transmitters.
Routine Syphilis Investigations
Public health departments in four states participated in ISP, a 5-year project involving collaborative efforts of public health and university researchers and community-based organizations. The project was conducted in Baton Rouge, LA, Columbia, SC, and Houston, TX for 4 years and for 1 year in Jackson, MS. All persons diagnosed with syphilis from January 1997 through June 1999 and reported to the local health departments in the reporting areas of the four research centers were eligible for this study. As part of standard local STD prevention activities, patients with early syphilis reported to the health department are contacted by specially trained local health department staff—Disease Intervention Specialists (DIS)—who confirm the diagnosis, gather information for the staging of syphilis, determine the infectious period, ascertain limited risk information, provide prevention counseling, and, depending on the stage of syphilis, elicit information on recent sex partners. DIS attempt to contact all possibly infected partners for examination and offer either curative or preventive treatment. At the end of each investigation, each partner is classified according to the following categories: “infected, treated,” “infected, not treated,” “previously treated,” “not infected,” “preventive treatment,” “preventive treatment refused,” and a group that includes partners who are not locatable for a local examination.
Enhanced Case Interview
We developed the Enhanced Case Interview (ECI) questionnaire to be administered by DIS to as many early syphilis patients as possible. The purpose of the questionnaire was: (1) to better describe persons with syphilis in communities with persisting syphilis in the 1990s; (2) to characterize persons who may sustain transmission in these communities; and (3) to assess potential priorities among persons with syphilis for partner notification and perhaps enhanced case-management based on characteristics that were associated with transmission of syphilis to partners.
To reach these goals, the ECI included questions about the person with syphilis as well as this person’s partner(s). We inquired about access to health care, about current education level and employment, and about sex partners and concurrency of partners during the previous 4 weeks. We asked about risk behaviors such as drug and alcohol use in the previous 3 months, drug treatment, incarceration, or detention in the previous year, and number of previous episodes of specific STD. We asked similar questions about the risk behaviors of the partner(s) of persons with syphilis, specifically partners in the previous 3 months.
In early 1997, one person trained all local DIS at each research center to administer the ECI questionnaire. Persons with primary, secondary, and early latent syphilis were interviewed. Since the questionnaire took 15 minutes to administer and DIS have multiple responsibilities, the addition of the ECI to the clinical encounter could interfere with normal activities. As a result, an ECI was not administered to all persons. Oral consent for the interview was obtained with the use of a standardized script. This study was approved by institutional review boards at each participating center and at the Centers for Disease Control and Prevention (CDC).
“Period partners” were defined as partners during the previous 3 months for persons with primary syphilis, during the previous 6 months for those with secondary syphilis, and during the previous year for those with early latent syphilis. “Previous month partners” were defined as partners during the previous month, regardless of the stage of syphilis. We defined a person as having concurrent partners if she/he had sex with two or more partners during 1 week in the previous month, and a person as having overlapping partners if she/he had period partners with overlapping dates of sexual exposure.
In order to provide information on persons who are likely to sustain syphilis transmission, a working definition of a “transmitter” was developed prior to data analyses. Since timing of infection suggests the directionality of transmission, we assumed that a person who has a period partner with syphilis at an earlier stage of disease than themselves is likely to have transmitted the infection to this period partner. Persons meeting this criterion were defined as transmitters and were compared to persons who also had at least one period partner.
We compared the reasons for seeking medical advice and the method of case detection of persons meeting the transmitter definition, and described the demographic and behavioral characteristics associated with being a transmitter. The behavioral characteristics of persons with syphilis associated with being a transmitter in univariate analysis with P < 0.1 were used in backward stepwise logistic regression to arrive at a final multivariate model. We added the variable number of partners to the final model, because of its possible role as a confounder. Interactions between variables in the final model were assessed by comparing maximum likelihood estimates with and without interaction terms. Data were entered locally, combined with routine interview data, and sent to CDC quarterly. Analyses were conducted using SAS (version 6.12, Cary, NC).
We also analyzed routinely collected syphilis case report data from persons with early syphilis who did not have an ECI during the same period in which the ECI was being administered from three research centers (Mississippi, South Carolina, and Texas). We determined whether having overlapping partners, or partners whose dates of sexual exposure overlapped, was a risk factor for being a transmitter in this group. There is no information in routinely collected syphilis case reports as to whether a partner had other partners, used drugs, or gave or received sex in exchange for money or drugs.
Patients With Early Syphilis
From January 1997 through June 1999, 1,908 patients with early syphilis received routine interviews and 743 (39%) also received the ECI. The number of persons with early syphilis and the proportion with completed ECIs varied by Center (Table 1). Compared with persons who did not receive the ECI, those who did were less likely to be more than 44 years old (P = 0.02), white (P = 0.02), or Hispanic (P = 0.002), more likely to have primary or secondary syphilis (P < 0.001), to be African American (P = 0.004), and to have at least one period partner (P < 0.001). There were no differences by gender.
Of the 743 persons with syphilis who had an ECI, slightly more than half were male, 627 (84%) were African American, and 304 (41%) had primary or secondary syphilis. Of the 373 with a partner in the previous month, only 33 (4.4%) were men who reported sex with men, and 6 (0.8%) were women who reported sex with women. Two hundred and fifty-five (35%) reported using alcohol daily, 147 (21%) marijuana daily, and 105 (15%) crack cocaine in the previous 3 months. Less than one percent of all persons with early syphilis reported injecting drugs or heroin or use of crystal, crank, or methamphetamine in the last 3 months.
The 743 persons with early syphilis named a total of 1,699 period partners (median 2), of whom 1,125 (66%, median 1) were located and examined. Of the 1,125 located period partners, 377 (33.5%) had syphilis, including 120 (10.7%) with primary or secondary syphilis. There were 296 (39.8%) persons who had at least one identified infected period partner.
There were 63 persons who were transmitters, i.e., had at least one infected partner at an earlier stage of syphilis. The proportion of patients with an ECI who met this definition varied by research center (Table 2). Among the persons who did not meet the definition of transmitter, 213 (36%) had sought care because of symptoms and 163 (27%) had been referred by a partner or the health department, whereas among transmitters only 7 (11%) had sought care because of symptoms but 45 (71%) had been referred (P < 0.001). This information was mirrored by the method of case detection; 67 (11%) nontransmitters and 29 (46%) transmitters had been identified through referral (P < 0.001). A similar pattern of seeking care and case detection was found at each site (data not shown).
The relationship between selected demographic and behavioral characteristics among those with at least one period partner and being a transmitter is shown in Table 3. Persons with two (odds ratio [OR] 5.8) or three or more partners (OR 6.0) in the previous month and persons who had concurrent partners (OR 2.3) in the previous month were more likely to be transmitters. Demographic variables, employment, education, drug use, drug treatment, incarceration, and previous episodes of STDs were not significantly associated with having a partner with an earlier stage of syphilis (Table 3). Persons with a partner who had two or more partners in the last 3 months (OR 2.1), a partner who had been in drug treatment (OR 3.7), or a partner who used marijuana (OR 2.0) were also more likely to be transmitters (Table 4).
The multivariate model demonstrating the independent contributions of risk characteristics for transmitters is shown in Table 5. Concurrent partners and having a partner who had been in drug treatment were independently associated with being a transmitter, although the confidence interval for concurrent partners slightly overlaps 1. There were no significant interactions between these variables. There was no additional independent contribution of having more than two partners in the last month.
Analysis of Data From Persons With Early Syphilis Without an ECI
There were 1,007 persons with early syphilis from Mississippi, South Carolina, and Texas who did not have an ECI. These patients named 1,921 sex partners, of whom 1,450 (75%) were located. Of these partners, 461 (24%) had syphilis. There were 319 who had at least one infected partner at an earlier stage. Among the 721 with at least one period partner, those who had one (OR 1.9, 95% CI 0.9–3.9) or two or more (OR 3.6, 95% CI 1.6–7.9) overlapping partners were more likely to be a transmitter.
This study characterized persons transmitting syphilis during a period of relatively low endemic rates. The results highlight the continued association between infection, transmission, and substance use. It also suggests the importance of concurrent or overlapping partners in syphilis transmission.
In these four research centers, most persons with syphilis were between the ages of 20 and 40, and about half of all persons with syphilis were women. African Americans continued to constitute a disproportionate number of persons with syphilis. These demographic characteristics are similar to those reported in a previous epidemiologic analysis performed during the last nationwide epidemic, 23 although rates of syphilis are now at historic lows.
We demonstrated that persons with syphilis who have concurrent or overlapping partners are at higher risk for meeting our definition of a transmitter. Although concurrency was most frequent among those with multiple partners, concurrency was an independant risk factor for being a transmitter, as we defined it in the ECI. Data from persons who did not have an ECI confirmed these findings. Concurrency of partners may enhance the transmission of syphilis for several reasons. Persons who have concurrent partners may have partners who also have concurrent partners. The multiplicity of partnerships among these persons with syphilis and their partners resembles the multiplicity of linkages found in dense sexual networks, which are known to efficiently transmit syphilis and other STDs. 18,20,21 Concurrent partnerships also allow a person, once infected by a partner, to spread the infection to a third person; this spread is less likely to occur if the person changes partners less frequently and the duration of partnerships is longer than the period of communicability of syphilis. In fact, the exchange of sex for goods, services, money, or drugs, 24 which has been repeatedly associated with STD transmission, may simply be a marker for a high number of concurrent partners. The importance of concurrent partners in the transmission dynamics of HIV and STDs has been illustrated by several mathematical models 14–17 and recently for chlamydia. 25
Prevention messages have traditionally emphasized reducing the number of sexual partners. While the number of recent sex partners did not emerge as an independent factor in predicting a transmitter of syphilis in multivariate analyses, most people with concurrent partners also had multiple partners. While the number of partners may be an important determinant of syphilis acquisition, we showed that having concurrent or overlapping partners are perhaps more important for syphilis transmission. The importance of this behavior (having concurrent partners) may be because of the increased likelihood of forming sexual networks. In addition to the traditional prevention message of reducing the total number of sex partners, prevention messages should also discourage having concurrent partnerships.
An interesting observation from this analysis is that there were differences in the number of transmitters identified at each center. Houston, a large city with the highest number of cases of early syphilis, had a significantly lower proportion of persons with early syphilis who met the transmitter definition. Baton Rouge and Columbia, which are smaller cities with similar rates of early syphilis, had higher proportions of persons with early syphilis meeting the transmitter definition. This observation raises some questions. Does the prevalence of concurrent partnerships play a more important role in areas with low rates of, or endemic, syphilis, rather than areas with high rates or hyperendemic syphilis? Does the prevalence of concurrent partnerships help determine the incidence of early syphilis in these areas?
Partner notification and referral during the epidemic of the early 1990s did not appear effective in controlling the epidemic 26; one weakness of partner notification was unlocatable partners. 26 In contrast, during a period of lower rates of early syphilis, we found that persons who were transmitters were often referred for care by partners or the health department. Although probably biased by our definition of a transmitter which required identification of partners, this finding suggests that sustained high quality partner notification is perhaps more important for syphilis elimination activities in foci of endemic syphilis than during large epidemics.
Our conclusions may be limited by the completeness of syphilis reporting. While syphilis is a reportable condition in all states, not all cases are detected or reported. However, the effect of incomplete detection and reporting would have a smaller impact in these research centers than in other areas because each center also had enhanced screening for syphilis in emergency rooms, jails, and delivery rooms, where additional cases were detected. 27 An additional limitation may be that less than half of all persons with early syphilis received an ECI. Persons who received an ECI had different demographic characteristics than those who did not. These differences may have influenced the risks we identified for being a presumptive transmitter, although none of the demographic characteristics were associated with being a transmitter. We cannot exclude the possibility that DIS used some conscious or unconscious criteria for whom to interview with the ECI. However, the importance of concurrent partners, redefined as overlapping partners, was reconfirmed by the analysis of contact tracing information from persons who did not have an ECI from three of the same research centers. This analysis suggested a “dose-response” relationship, with an increasing odds ratio as the number of overlapping partners increased, with confidence intervals that did not include 1.
The importance of concurrency among persons with syphilis and their partners emerged despite limitations in our definition of a transmitter. Contacts may have acquired an early stage of syphilis from a person other than the transmitter we defined in the analysis. Since persons with one or more partners with an earlier stage of syphilis were defined as transmitters, persons with primary (and highly infectious) syphilis could not be classified as transmitters. Our definition also most likely excluded other persons who transmitted syphilis, because 25% to 33% of partners were not located. The inability to find all partners and ascertain their infection status and the possibility that contacts acquired syphilis from another person both would tend to introduce a misclassification bias that would decrease the estimated importance of concurrent partners. Unlocatable and not located partners such as commercial sex workers or very casual sex partners tend to be elusive to public health activities. 3,20
There are important implications of our findings for syphilis control and elimination efforts. Persons with syphilis and concurrent/overlapping partners might be given enhanced case management, such as “cluster interviews”22 or other interventions. 28 Some type of preventive case management may also be beneficial for persons who do not have syphilis but who report such risk behaviors and are linked through sexual networks to persons who have syphilis. Currently the syphilis interview does not routinely ascertain behavioral risk factors. The identification and recording of such behaviors, such as illicit substance use, should be considered as an important possible addition to routine interviews and locally collected data elements. Information on overlapping partners is currently not easily extractable from electronically collected data. Existing software could be modified to identify persons with overlapping partners and to show networks. Additionally, persons performing the interviews can easily note whether any partners overlap during elicitation of dates of exposure to partners. Use of this data could help focus activities in the remaining areas of the United States with syphilis.
1. Primary and Secondary Syphilis–United States, 1998. MMWR Morb Mortal Wkly Rpt 1999; 48:873–878.
2. Division of STD Prevention. Sexually Transmitted Disease Surveillance, 1998. US Department of Health and Human Services, Public Health Service. Atlanta: Centers for Disease Control and Prevention, September 1999.
3. Gunn RA, Montes JM, Toomey KE, et al. Syphilis in San Diego County 1983–1992: crack cocaine, prostitution, and the limitations of partner notification. Sex Transm Dis 1995; 22: 60–66.
4. Fleming DT, Wasserheit JN. From epidemiologic synergy to public health policy and practice: the contribution of other sexually transmitted diseases to sexual transmission of HIV infection. Sex Transm Infect 1999; 75: 3–17.
5. Koumans EH, Sternberg M, Gwinn M, et al. Geographic variation of HIV infection in childbearing women with syphilis in the United States. AIDS 2000; 14: 279–287.
6. Blocker ME, Levine WC, St Louis ME. HIV prevalence in patients with syphilis, United States. Sex Transm Dis 2000; 27: 53–59.
7. Edlin BR, Irwin KL, Faruque S, et al. Intersecting epidemics–crack cocaine use and HIV infection among inner-city young adults. N Engl J Med 1994; 331: 1422–1427.
8. Ellerbrock TV, Lieb S, Harrington PE, et al. Heterosexually transmitted human immunodeficiency virus infection among pregnant women in a rural Florida community. New Engl J Med 1992; 327: 1704–1709.
9. Nakashima AK, Rolfs RT, Flock ML, et al. Epidemiology of syphilis in the United States, 1941–1993. Sex Transm Dis 1996; 23: 16–23.
10. Garnett GP, Aral SO, Hoyle DV, et al. The natural history of syphilis. Implications for the transmission dynamics and control of infection. Sex Transm Dis 1997; 24: 185–200.
11. Oxman GL, Smolkowski K, Noell J. Mathematical modeling of epidemic syphilis transmission, implications for syphilis control programs. Sex Transm Dis 1996; 23: 30–39.
12. Aral SO. The social context of syphilis persistence in the southeastern United States. Sex Transm Dis 1996; 23: 9–15.
13. Wasserheit JN, Aral SO. The dynamic topology of sexually transmitted disease epidemics: implications for prevention strategies. J Infect Dis 1996; 174 (Suppl 2): S201–213
14. Watts CH, May RM. The influence of concurrent partnerships on the dynamics of HIV/AIDS. Math Biosci 1992; 108: 89–104.
15. Morris M, Kretzschmar M. Concurrent partnerships and transmission dynamics in networks. Soc Netw 1995; 17: 299–319.
16. Morris M, Kretzschmar M. Concurrent partnerships and the spread of HIV. AIDS 1997; 11: 641–648.
17. Ghani AC, Garnett GP. Risks of acquiring and transmitting sexually transmitted diseases in sexual partner networks. Sex Transm Dis 2000; 27: 579–587.
18. Engelgau MM, Woernle CH, Rolfs RT, et al. Control of epidemic early syphilis: the results of an intervention campaign using social networks. Sex Transm Dis 1995; 22: 203–209.
19. Rosenberg D, Moseley K, Kahn R, et al. Networks of persons with syphilis and at risk for syphilis in Louisiana: evidence of core transmitters. Sex Transm Dis 1999; 26: 108–114.
20. Woodhouse DE, Rothenberg RB, Potterat JJ, et al. Mapping a social network of heterosexuals at high risk for HIV infection. AIDS 1994; 8: 1331–1336.
21. Yorke JA, Hethcote HW, Nold A. Dynamics and control of the transmission of gonorrhea. Sex Transm Dis 1978; 5: 51–56.
22. Centers for Disease Control and Prevention, Division of STD Prevention. Program Operation Guidelines for STD prevention, 2000.
23. Rolfs RT, Nakashima AK. Epidemiology of primary and secondary syphilis in the United States, 1981 through 1989. JAMA 1990; 264: 1432–1437.
24. Baseman J, Ross MW, Williams M. Sale of sex for drugs and drugs for sex: an economic context of sexual risk behavior for STDs. Sex Transm Dis 1999; 26: 444–449.
25. Potterat JJ, Zimmerman-Rogers H, Muth SQ, et al. Chlamydia transmission: concurrency, reproduction number, and the epidemic trajectory. Am J Epidemiol 1999; 150: 1331–1339.
26. Andrus JK, Fleming DW, Harger DR, et al. Partner notification: can it control epidemic syphilis? Ann Intern Med 1990; 112: 481–483.
27. Finelli L, Farley TP, Gibson JJ, et al. Prevalence monitoring in syphilis surveillance: results from a multicenter research program. In: Abstracts of the 1998 National STD Prevention Conference; December 6–9 1998; Dallas, TX. Abstract A6:91.
28. Rothenberg R, Kimbrough L, Lewis-Hardy R, et al. Social network methods for endemic foci of syphilis. Sex Transm Dis 2000; 27: 12–18.