Samoff, Erika PhD, MPH*; Koumans, Emilia H. MD, MPH*; Katkowsky, Steven MD†; Shouse, R Luke MD, MPH‡; Markowitz, Lauri E. MD*; The Fulton County Disease Investigation Working Group
From the *Division of Sexually Transmitted Disease Prevention, National Center for HIV, STDs, and Tuberculosis (TB) (NCHSTP), Centers for Disease Control and Prevention (CDC), Atlanta, Georgia; the †Fulton County District Health Office and Department of Health and Wellness; and the ‡HIV/STD Epidemiology Branch, Division of Public Health, Georgia Department of Human Resources, Atlanta, Georgia
The authors acknowledge the work of the Centers for Disease Control Rapid Response Team working in Fulton and DeKalb Counties in 2003, which informed this manuscript; the assistance of Jamie Howgate and Deborah McElroy; and Matthew Hogben, Richard Kahn, and Tom Peterman for their comments on the manuscript.
This study was supported by funds from the State of Georgia Department of Human Resources, the Fulton County Department of Health and Wellness, and the Division of Sexually Transmitted Disease Prevention and Syphilis Elimination Effort, Centers for Disease Control and Prevention (Atlanta, Georgia).
The Fulton County Disease Investigation Working Group was composed of: Michelle L. Allen, BA, Reginald Brown, Elijah Cameron, MA, Donald Chastang, BS, Ruby Lewis-Hardy, BA, and Pradnya Tambe, MD.
Correspondence: Erika Samoff, PhD, MPH, Department of Obstetrics and Gynecology, University of California, San Francisco, 850 Marina Bay Pkwy, Bldg. P, Richmond CA 94804. E-mail: email@example.com.
Received for publication February 28, 2006, and accepted October 17, 2006.
CONTACT TRACING, THE LOCATING, INTERVIEWING, and treatment of sex partners of persons with a sexually transmitted disease (STD), has historically been regarded as an important control measure for syphilis1 as well as other sexually transmitted diseases. Contact tracing is only effective, however, to the extent that infected persons are able and willing to give locating information for sex partners. Because locating information for anonymous or casual sexual partners may be more difficult to obtain than for long-term partners, contact-tracing activities may be less effective in populations characterized by a high proportion of anonymous or casual partners.
Studies evaluating patients with syphilis or STD clinic attendees have reported anonymous or casual sex partnerships among men reporting sex with men (MSM).2–4 These data support the common impression that contact-tracing activities locate fewer partners among MSM than among men reporting sex with women only (MSWO); however, only one published statistical analysis supports this conclusion.5 Other studies report similar numbers of partners located and new syphilis cases diagnosed among MSM and MSWO STD clinic attendees.3,6
Rates of syphilis among heterosexuals in the southern United States have historically been higher than rates among heterosexuals in other areas of the country. In recent years, the proportion of syphilis cases that occur among MSM has increased dramatically in the South like in other areas,8 resulting in a population with relatively high rates of syphilis among both MSM and MSWO. This disease distribution allows investigation of contact-tracing outcomes among MSM and MSWO syphilis index patients investigated by the same personnel. To evaluate whether syphilis contact-tracing outcomes differ by sexual behavior in a large city in the South, we compared contact-tracing outcomes among 2003 MSM and MSWO patients with syphilis from Fulton County, Georgia.
Syphilis Case Investigation
The Fulton County Department of Health and Wellness (FDHW) Communicable Disease Prevention Branch (in conjunction with the State of Georgia Syphilis Elimination Project) routinely collects data on reportable infectious diseases, including syphilis. When a case of syphilis is diagnosed by private or public healthcare providers, the state of Georgia requires submission of a case report including basic demographics and diagnosis information to the Georgia Division of Public Health. After data entry and prioritization of follow up, the state notifies county health departments of cases for which patient follow up is needed. Standard procedures performed by trained FDHW staff include a patient interview during which information about disease progression, STD history, and sexual behaviors are elicited. To identify additional persons at risk for syphilis, the index patient is asked to report the number of sexual partners during the period when he or she had (potentially) infectious syphilis and to give names and locating information for these partners. The infectious period is determined by the syphilis stage: 3 months before diagnosis for primary syphilis, 6 months before diagnosis for secondary syphilis, and 1 year before diagnosis for early-latent syphilis. The patient is also asked to name friends or associates whom the index patient believes are at high risk of syphilis; this step is not practiced in all local areas but is standard in Fulton County. FDHW staff attempt to locate and interview all contacts (sex partners and nonsexual contacts) named by the index patient and to bring them to the clinic for evaluation and treatment if appropriate. Contacts may be uninfected, diagnosed with syphilis and treated, or prophylactically treated if infection is possible but not diagnosable (i.e., exposed to an index patient but showing no sign of infection and in the “window period” when an immune response to infection may not yet be detectable).
In January 2004, 762 cases of syphilis were reported by Fulton County; of these, 597 were among men. We reviewed all available (547) health department records of early (primary, secondary, or early-latent) syphilis cases in male Fulton County residents; these records made up 92% of the 597 early syphilis cases among men reported by Fulton County for 2003. Although the majority of these records included interview data, a subset of these cases (146 [27%]) did not as a result of insufficient or incorrect locating data or refusal of the interview by the index patient. Because behavioral and contact data are collected during the interview process, comparative analyses were limited to cases with interview data.
Data on index patient characteristics from the national syphilis case report form (73.54) and from the extended interview used in Fulton County were evaluated to assess demographics, provider and disease characteristics, and sexual behaviors. Healthcare provider, race, and stage information were obtained from the national case report form.
To assess MSM status, data from 3 sources were evaluated: the question “have you had sex with a man since 1978?” from the national case report form; the question “what was the gender of your sex partners in the past 12 months?” with the response categories “male,” “female,” and “both” from the extended interview form; and the recorded gender of sex partners named during the interview with FDHW staff. If any of these variables indicated MSM activity, the index patient was considered a MSM. Sexual behavior data, including reported anonymous partnerships in the 3 months before the interview, were abstracted from the extended interview forms. The number of total sex partners reported during the infectious period was abstracted from the national case report form.
Records of contact investigations (interviews of all sex partners and nonsexual contacts named by the index patient during the interview) were merged with these data to generate a database in which index patients and their contacts were linked. Data were available for those contacts investigated by a FDHW staff member; contact type (sex partner or nonsexual contact) was indicated in the contact record; all contacts for whom an investigation was initiated were included in the analysis. The number of unnamed infectious period sex partners was calculated by subtracting the number of named sex partner contacts from the total number of infectious period sex partners.
To assess outcomes of contact tracing, we compared the numbers per case of contacts investigated, located, and diagnosed with syphilis; evaluated the number and proportion of contacts who were prophylactically treated for syphilis; and evaluated the stage of disease and treatment history of infected contacts. These outcomes were compared by behavioral sexual orientation. Numbers of contacts per case were not normally distributed. Because a subset of syphilis cases had no contact information, medians were often 0 or 1 and did not effectively represent the distribution of values. Therefore, means were presented in the text. As is appropriate for nonnormally distributed values, the nonparametric Wilcoxon rank sum statistic was used for comparisons. The χ2 statistic was used to evaluate differences between proportions.
We reviewed 547 reports of early syphilis cases in male residents of Fulton County, Georgia, that occurred in 2003. After exclusion of cases without interview data, 401 (73%) cases remained. Of these 401 index patients, 243 (61%) reported MSM behavior and 158 (39%) reported no MSM behavior. Population characteristics are presented in Table 1. In comparison to MSWO index patients, a higher proportion of MSM index patients were white, were diagnosed by private providers, were diagnosed with infectious (primary or secondary) syphilis, and reported anonymous sex acts in the past 3 months (all P <0.01).
During the interview process, the index patient is asked to indicate the number of partners with whom he has had sex and is then asked to name both these partners and nonsexual contacts; those for whom names are given are called “contacts.” Contacts may be either sexual partners or nonsexual contacts. In aggregate, MSM index patients reported having 764 total sex partners during the infectious period and named 191 (25%); MSWO reported 387 total sex partners and named 92 (24%). The total 207 contacts to MSM investigated included these 191 named sexual contacts as well as 16 named nonsexual contacts; the total 116 contacts to MSWO investigated included 92 named sex partners and 24 named nonsexual contacts.
Data on the number of contacts named (name and locating information provided), located, and diagnosed with syphilis per case are presented in Table 2. Interviews of MSM index patients resulted in higher numbers of contacts named (mean 0.80 vs. 0.72, P = 0.04) and located (mean 0.65 vs. 0.57, P = 0.01) per case than interviews of MSWO index patients. There were no significant differences in the number of contacts of MSM and MSWO index patients diagnosed with syphilis per case (P = 0.46).
A smaller proportion of MSWO index patients than MSM index patients were willing or able to name contacts, resulting in lower contact-tracing yield among this population. Among MSM index patients, 136 of 243 (56%) named at least one contact, whereas among MSWO, 71 of 158 (45%) named at least one contact (P = 0.03). Among those index patients willing to name at least one contact, there were no significant differences between the numbers of contacts per case named, located, and diagnosed with syphilis.
Prevention and Case-Finding Outcomes
The number of uninfected contacts and the proportion of contacts receiving prophylactic treatment are presented in Table 2. Neither the number of uninfected contacts nor the number of contacts per case prophylactically treated (0.23 for MSM, 0.17 for MSWO) differed significantly.
The number of infected contacts, the proportion with infectious syphilis, and the proportion treated resulting from contact-tracing efforts are also presented in Table 2. The proportion of contacts with infectious (primary and secondary) syphilis was higher for MSM index patients, although this difference was not significant. We evaluated the proportion of infected contacts who were newly diagnosed with syphilis (as opposed to diagnosed and treated by another clinician before the contact-tracing process). For MSM index patients, the contact tracing process located 0.10 new syphilis cases per case-patient; for MSWO, the contact-tracing process located 0.14 new syphilis cases per case patient; the differences between these numbers was not significant. The majority of new cases were contacts to case-patients identified with early-latent disease (50% and 63% of newly diagnosed contacts to MSM and MSWO, respectively, were contacts of case-patients with early-latent disease).
Factors Affecting Contact Notification Yield
Stage of Disease.
A higher proportion of MSM were diagnosed with infectious syphilis than MSWO. Neither the number of contacts located (mean 0.67 for primary, 0.62 for secondary, and 0.62 for early-latent syphilis; P = 0.82) nor the number of infected contacts (mean 0.24 for primary, 0.23 for secondary, and 0.23 for early-latent; P = 0.99) per case differed by stage at diagnosis of the index patient; therefore, stage at diagnosis does not explain differences in contact-tracing efficacy.
A higher proportion of MSM (45%) reported anonymous sex acts in the past 3 months than MSWO (32%, P <0.01). However, these groups had approximately similar proportions of unnamed sex partners. In aggregate, MSM index patients reported 764 sex partners and named 191 of these partners (25%); MSWO reported 387 total partners and named 92 (24%). Although the mean number of unnamed sex partners per case was higher for MSM (2.4) than MSWO (1.9) index patients, this difference was nonsignificant (P = 0.39).
The effect of increasing numbers of MSM index patients on syphilis case follow up is much discussed, although few studies provide statistical analyses to fuel these discussions. Our study comparing case finding by the same trained health department personnel demonstrates that outcomes of contact tracing in a large city in the southern United States are similar for MSM and MSWO. Although the success of contact-tracing efforts is likely to vary with differing local context, these data do not support the assumption that contact-tracing efforts will invariably be less effective among MSM.
In Fulton County and elsewhere,2–4 higher proportions of MSM index patients report anonymous partners than MSWO index patients. It is logical to expect that differences in contact-tracing outcomes would result. However, if MSM index patients name similar numbers of contacts as MSWO index patients, differences in contact-tracing outcomes would not be reflected in summaries of contacts located or diagnosed with syphilis. Rather, this difference may appear as a higher number of unnamed sex partners.
The work of Andrus et al9 evaluates differences in unnamed partners between syphilis and gonorrhea case-patients and proposes that if the proportion of sex partners who are not named is high, an epidemic cannot be controlled by contact tracing. The proportion of partners not named is not commonly reported in contact-tracing analyses; however, this may be a key determinant for evaluating the role that contact tracing should play in an epidemic control effort using multiple strategies and may be an effective measurement of what differentiates populations of MSM and MSWO index patients. It is important to note that this outcome may not differ between MSM and MSWO as was the case in Fulton County in 2003; approximately 75% of sex partners were not named for both groups. Similarly, it may be valuable to compare the numbers of partners not named, strikingly different for the 2 groups in Fulton County (573 for MSM, 297 for MSWO). An assessment of numbers or proportions of unnamed partners should be considered when evaluating the contribution of contact tracing programs to STD prevention.
Mistrust of health department investigators, personal feelings of efficacy in partner notification, and a high proportion of index patients receiving care from private providers have been among the reasons suggested for decreased contact-tracing efficacy among MSM,10 and these factors may have affected contact-tracing efforts in Fulton County. A number of other factors may facilitate contact-tracing efficacy among MSM index patients. A high proportion of cases among MSM reported in Fulton County were diagnosed at the public STD clinic, facilitating follow up and investigation of the index patient. On recognition of increasing MSM cases in Fulton County, efforts were made to improve follow up of MSM index patients, which has resulted in improved relations with healthcare providers working with the MSM community, improved knowledge of community norms, and increased interaction with MSM community-based organizations.
We present numbers of contacts for each outcome to allow explicit comparison of the outcomes of health department investigations. A comparison of numbers found per disease investigation specialist (DIS) investigation time period or other measurement of resources expended (as performed by Peterman and colleagues in a randomized, controlled comparison of partner-tracing methods3) would provide a clearer picture of the value of contact tracing to control of syphilis transmission and would provide valuable information to decision-makers at the national and local level.
This analysis was subject to certain limitations. Contact-tracing efficacy as measured in this analysis was lower than that found for an earlier investigation involving Fulton County personnel,11 perhaps because these investigations were performed before rather than as part of the outbreak response. The epidemiology of syphilis in the southern United States differs historically from other regions of the country, and differences in the historical presence of this disease in the community and in community prevalence may limit the generalizability of these findings. However, because the southern United States accounts for more syphilis cases than any other region in the United States, these data are valuable for nationwide syphilis control efforts.
The proportion of syphilis cases characterized by MSM behavior has increased in many large cities over the past 5 years.8,12–14 It is possible that men coded as MSWO in this analysis may in fact be MSM, although given the largely similar patterns of contact tracing, this would not be likely to change the conclusions of the analysis. Results from Fulton County, where the prevalence of syphilis is high among both men and women, may not be generalizable to patient populations composed almost entirely of MSM. It is possible that the success of contact-tracing activities differs in these cities; one review suggests contact-tracing yields were lower in locations where most index patients reported MSM behavior.7 The number of reported partners among MSM in this community is lower than reported in some other communities.
The majority of interviewed index patients were diagnosed at the Fulton County STD clinic. Data from this analysis suggest that patients seeing private providers are less likely to name contacts; it is possible that patients seeing private providers were less willing to be interviewed, and that a selection bias was introduced in this manner. No data are available to evaluate this possibility.
Finally, a high proportion of case-patients were not located for interview. It is possible that these case-patients were more likely to be MSM; not including the unlocated contacts of these case-patients may lead to the appearance that contact-tracing yield is as high among MSM as among MSWO. It is not possible to assess this question with the data available to us.
This analysis supports continued efforts to perform contact-tracing activities among MSM as well as MSWO syphilis index patients. Further analyses of contact tracing among MSM index patients with different population characteristics will contribute to the ongoing assessment of contact-tracing efficacy. Studies modeling the effect of unnamed partners may help to evaluate and better define the potential contributions and roles of contact tracing and other syphilis control strategies.
1. Cates WJ, Rothenberg RB, Blount JH. Syphilis control: The historic context and epidemiologic basis for interrupting sexual transmission of Treponema pallidum
. Sex Transm Dis 1996; 23:68–75.
2. Kim AA, Kent C, McFarland W, Klausner JD. Cruising on the Internet highway. J Acquir Immune Defic Syndr 2001; 28:89–93.
3. Peterman TA, Toomey KE, Dicker LW, et al. Partner notification for syphilis: A randomized, controlled trial of three approaches. Sex Transm Dis 1997; 24:511–518.
4. Taylor M, Ayanalem G, Smith L, 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.
5. Rogstad KE, Clementson C, Ahmed-Jushuf IH. Contact tracing for gonorrhoea in homosexual and heterosexual men. Int J STD AIDS 1999; 10:536–538.
6. Jayaraman GC, Read RR, Singh A. Characteristics of individuals with male-to-male and heterosexually acquired infectious syphilis during an outbreak in Calgary, Alberta, Canada. Sex Transm Dis 2003; 30:315–319.
7. Brewer DD. Case-finding effectiveness of partner notification and cluster investigation for sexually transmitted diseases/HIV. Sex Transm Dis 2005; 32:78–83.
8. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance, 2002. U.S. Department of Health and Human Services, 2003.
9. Andrus JK, Fleming DW, Harger DR, et al. Partner notification: Can it control epidemic syphilis? Ann Intern Med 1990; 112:539–543.
10. Hogben M, Paffel J, Broussard D, et al. Syphilis partner notification with men who have sex with men: A review and commentary. Sex Transm Dis 2005; 32:S43–S47.
11. Rothenberg R, Kimbrough L, Lewis-Hardy R, et al. Social network methods for endemic foci of syphilis: A pilot project. Sex Transm Dis 2000; 27:12–18.
12. Centers for Disease Control and Prevention. Outbreak of syphilis among men who have sex with men—Southern California, 2000. MMWR Morb Mortal Wkly Rep 2001; 50:117–120.
13. Centers for Disease Control and Prevention. Primary and secondary syphilis among men who have sex with men—New York City, 2001. MMWR Morb Mortal Wkly Rep 2001; 51:853–856.
14. Centers for Disease Control and Prevention. Outbreak of syphilis among men who have sex with men—Southern California, 2000. MMWR Morb Mortal Wkly Rep 2001; 50:117–120.