Only 6 gonorrhea clusters were classified as outbreak areas, based on a duration less than 5 years (Fig. 2). Two outbreak areas occurred in urban areas and 1 in a micropolitan area. The remainder (n = 3) were geographically large outbreaks capturing combinations of rural, small town, micropolitan, and urban areas (Fig. 2). The RRs for core areas (median RR, 3.4) were significantly greater than the RRs for outbreak areas (median RR, 2.1) of gonorrhea (P < 0.01).
Syphilis Core Areas and Outbreaks
Overall, average syphilis rates were low in the western mountains and higher in the central piedmont and eastern parts of the state, with particularly high rates in the southern rural part of the state (Fig. 1).
Seventeen clusters of elevated syphilis rates were identified across North Carolina, concentrated predominantly in the central piedmont region (Fig. 2). Ten clusters met the criteria for core areas, with 80% (n = 8) in urban centers; 20% (n = 2) were geographically large enough to capture combinations of rural, small town, micropolitan, and urban areas. The large size and diversity of these remaining 2 core areas suggest the possibility of dynamic migratory outbreaks, rather than true core areas. Again, no core areas occurred exclusively in small towns or rural areas.
Seven syphilis clusters were identified as outbreaks (Fig. 2). Three outbreaks occurred in urban areas, 1 in a micropolitan area, 1 in a small town, and 2 outbreaks that were extremely large capturing combinations of rural, small town, micropolitan, and urban areas (Fig. 2). No outbreaks occurred in rural areas. Although the RRs for outbreak areas (median RR, 10.1) were noticeably greater than the RRs for core areas (median RR, 5.3) of syphilis, the difference was not statistically significant.
Twice as many gonorrhea core areas (n = 20) were identified compared with syphilis core areas (n = 10). At the same time, all 10 (100%) of the syphilis core areas overlapped with gonorrhea core areas (50%; Fig. 3). Seven of the 10 overlapping core areas were in urban areas. The remaining 3 overlapping core areas included rural, small town, micropolitan, and urban environments (Fig. 3). The remaining 10 gonorrhea core areas that did not have matching overlapping syphilis core areas were located in the urban centers of otherwise rural areas (Fig. 3).
For the most part, gonorrhea and syphilis outbreaks did not coexist (Fig. 3). In 1 case, the outer edge of a geographically large gonorrhea outbreak overlapped with a geographically large syphilis outbreak, and in another case, the outer edge of a geographically large gonorrhea outbreak captured a small syphilis outbreak. However, in most cases, outbreak areas were adjacent to each other and did not touch.
The length of time clusters persisted for gonorrhea and syphilis was bimodal. That is, clusters were either short (≤3 consecutive years) or the longest (full) period allowed by SaTScan (90% of period: 5 consecutive years for gonorrhea and 10.5 consecutive years for syphilis). These 2 periods (≤3 years, or the full period) accounted for 85% of gonorrhea clusters and 82% of syphilis clusters. Changing the threshold for outbreaks and core areas anywhere between 3 and 10 years did not significantly change the results for syphilis. Lowering the threshold from 5 to 4 or 3 years does not significantly change the core area results for gonorrhea but does decrease the number of outbreaks by half. We could not see the effect of raising the threshold for gonorrhea because there were only 6 years of data available, and SaTSCan allows a maximal period of 90% of the total time available for a given cluster.
We qualitatively assessed the spatial correlation between gonorrhea clusters and rurality by overlaying the disease clusters on percent rural instead of RUCA and found that the core areas for both gonorrhea and syphilis still occurred primarily in urban (<20% rural) areas, with the exception of 2 geographically large, overlapping core areas that captured combinations of rural, small town, micropolitan, and urban areas (Fig. 3).
As expected, a greater number of statistically significant clusters were identified for both STIs when the analyses were stratified by region (Fig. 4). However, despite stratification, no clusters were identifiable in rural areas. Although no syphilis clusters were identified in the mountain region prior to stratification (Fig. 3), 1 cluster was identified in an urban area of the mountain region after stratification (Fig. 4).
Gonorrhea and syphilis rates were high for rural parts of North Carolina (Fig. 1); however, most core areas for both gonorrhea and syphilis still occurred in urban areas (Figs. 2–4). No core areas occurred exclusively in small towns or rural areas. When core areas included rural areas and small towns, they were large enough to include neighboring micropolitan and urban environments. These observations, independently and by different analytic methods, support previous findings that there can be bridging of infection from core areas to other areas by travel.16
We hypothesize that the communities of rural North Carolina are too small and isolated for STI epidemics to persist at endemic levels and thus create core areas. As a result, the main pathway of STI transmission may be through the interconnectedness of urban, micropolitan, small town, and rural areas. Core groups may exist in rural areas, with sexual network dynamics conducive to STI transmission, but this core group on its own may not be big enough to act as a reservoir of persistent infection. Consequently, infection in rural areas may be driven by core group connections to high rate core areas in nearby urban centers, micropolitans, or small towns or combinations thereof.
For North Carolina, urban core areas may have a more continuous magnifying effect on rural rates in the coastal region, but a discontinuous or intermittent effect on rural rates in the mountain region. This is particularly evident in the mountains where rural rates are low, although urban cores areas were identified in the region (Fig. 4). Rural-urban commuting area classification considers rural-urban travel commuting times; however, the difference in the rural-urban connection between mountain and coastal regions is likely influenced by additional factors. For instance, the mountain region may have a more centralized urban-rural connectivity, where the mountain topography causes each rural area to be primarily connected to only one “down valley” urban core area, whereas in the coastal region, the flat coastal topography may facilitate rural areas being connected to multiple surrounding urban core areas. A public health implication of this finding is that controlling STIs in one versus multiple urban core areas may need to be a critical consideration to reducing the high STI rates, especially in rural areas that are highly connected to multiple surrounding urban areas such as in the coastal region of North Carolina.
We had 12 years of syphilis data (1999–2010) and only 6 years of gonorrhea data (2005–2010). We used 5 consecutive years as the threshold for differentiating core from outbreak for both syphilis and gonorrhea, based on previous syphilis studies in urban areas8,9 and given the lack of studies on gonorrhea appropriate thresholds. We still found twice as many gonorrhea core areas as syphilis core areas, despite the shorter duration of gonorrhea data and syphilis-based thresholds. The greater number of gonorrhea core areas may be caused by the higher and more widespread incidence and prevalence of gonorrhea compared with syphilis.
It is still questionable whether or not 5 years is the optimum threshold for differentiating core areas from outbreaks for gonorrhea. Gonorrhea is a faster cycling disease than syphilis, with a shorter time to reinfection, which may mean that the 5-year threshold is too long for gonorrhea. At the same time, we found more core areas than outbreaks using the 5-year threshold for gonorrhea.
All syphilis core areas fell within, or overlapped with, gonorrhea core areas. At the same time, we know of at least one syphilis outbreak of long duration41 that met the criteria for core and so may be considered misclassified. This observation demonstrates the difficultly that can occur when differentiating outbreak areas from core areas.
Another significant limitation to our approach to classifying core versus outbreak clusters is that it is not possible to identify outbreaks that occur within core areas. That is, a core and an outbreak cannot be identified for the same place using the SaTScan detection method. The only way an outbreak can be seen within a core is if the analysis is limited to the boundaries of a core area. Two alternative methods would be to compliment the cluster detection and core/outbreak analysis with a sexual network analysis41 or an analysis of the rate of change within core areas.
In a previous study, we compared the characteristics of clusters on a very local level8 and found that the characteristics of cases from the core were very different from outbreak areas but very similar to noncore and nonoutbreak areas. Ideally, one would conduct a similar analysis and characterize the sociocultural demographics of core and outbreak clusters identified across North Carolina and then look for sociodemographic and correlations that could lend insight and context into the observed spatial patterns. However, the limits of surveillance data, combined with North Carolina State Health Department methods to protect case confidentiality, severely limited the gonorrhea and syphilis data we received, so we could not characterize clusters. We could summarize the neighborhood characteristics of core areas, in aggregate for the state; however, we would be at high risk for falling into the trap of ecologic fallacy without the individual-level case characteristics to provide context and aid interpretation.
Our methods may be of use in routine surveillance and outbreak detection of both gonorrhea and syphilis infections and may help focus more targeted surveillance, contact tracing, and cases follow-up. Modern data analysis packages are being used to write programs that import surveillance data into SaTScan and routinely run spatial and temporal analyses to detect outbreaks regularly, for instance, every month or every quarter, for other infectious diseases already.42
Routine spatiotemporal surveillance and analysis is particularly important for STIs because the core areas we identified for gonorrhea and syphilis are not isolated, single-jurisdiction pockets of infection. Rather, they involve rural-urban communities, therefore requiring synchronized management to address infections, which can bounce back and forth between cases of different jurisdictions (ping-pong infections). Multi-jurisdiction coordination may be particularly important for both “old” infections such as syphilis43 and “new” infections such as lymphogranuloma venereum,44 where infection is maintained and controlled in an urban area, but when combined with the long distances between source contacts and cases for these infections, these previously well-controlled infections may rise and spread to new and naive areas.
From a rural public health practice perspective, even if a core group of transmitters exists locally, STI rates in the community may still be largely driven by the interconnectedness of the sexual network to, and STI rates in, surrounding urban and micropolitan areas. From an urban public health practice perspective, STI dynamics in core areas may be impacting STI rates both locally and regionally, influencing, maybe even driving, STI rates in neighboring rural and small town communities. Consequently, directly addressing STIs in urban and micropolitan communities may also indirectly help address STI rates in rural and small town communities, presenting opportunities for mutually beneficial collaborations between communities.
The general absence of core areas of infection for gonorrhea and syphilis in rural environments supports the hypothesis of interconnected sexual networks between rural, micropolitan, and urban core groups. Consequently, bridging of infection from adjacent, peripheral, or distant core areas by travel of core group members may be an important mode of STI entry and transmission for rural areas. Interjurisdiction coordination of STI prevention and treatment activities may be necessary to control STIs in rural environments.
Appendix A SaTScan parameterization for cluster detection
Time precision: yearly. Note: we wanted the analysis by quarter so each quarter was assigned a year before bringing the data into SaTScan.
Type of analysis: retrospective, space-time
Probability model: Poisson
Scan for areas with: high rates
Time aggregation: quarter
Length: 1 quarter
MC replications: 999
Advanced analysis features:
- Maximum spatial cluster size: 5%
- Spatial window shape: circular
- Temporal window: 90% of the study period (maximum allowed)
1. Rothenberg RB. The geography of gonorrhea. Empirical demonstration of core group transmission. Am J Epidemiol 1983; 117: 688–694.
2. Yorke JA, Hethcote HW, Nold A. Dynamics and control of the transmission of gonorrhea. Sex Transm Dis 1978; 5: 51–56.
3. Alvarez-Dardet C, Marquez S, Perea EJ. Urban clusters of sexually transmitted diseases in the city of Seville, Spain. Sex Transm Dis 1985; 12: 166–168.
4. Becker KM, Glass GE, Brathwaite W, et al.. Geographic epidemiology of gonorrhea in Baltimore, Maryland, using a geographic information system. Am J Epidemiol 1998; 147: 709–716.
5. Bernstein KT, Curriero FC, Jennings JM, et al.. Defining core gonorrhea transmission utilizing spatial data. Am J Epidemiol 2004; 160: 51–58.
6. Bush KR, Henderson EA, Dunn J, et al.. Mapping the core: Chlamydia and gonorrhea infections in Calgary, Alberta. Sex Transm Dis 2008; 35: 291–297.
7. Ellen JM, Hessol NA, Kohn RP, et al.. An investigation of geographic clustering of repeat cases of gonorrhea and chlamydial infection in San Francisco, 1989–1993: Evidence for core groups. J Infect Dis 1997; 175: 1519–1522.
8. Gesink DC, Sullivan AB, Miller WC, et al.. Sexually transmitted disease core theory: Roles of person, place, and time. Am J Epidemiol 2011; 174: 81–89.
9. Gesink Law DC, Bernstein KT, Serre ML, et al.. Modeling a syphilis outbreak through space and time using the Bayesian maximum entropy approach. Ann Epidemiol 2006; 16: 797–804.
10. Jolly AM, Wylie JL. Gonorrhoea and chlamydia core groups and sexual networks in Manitoba. Sex Transm Infect 2002; 78 (suppl 1): i145–i151.
11. Law DC, Serre ML, Christakos G, et al.. Spatial analysis and mapping of sexually transmitted diseases to optimise intervention and prevention strategies. Sex Transm Infect 2004; 80: 294–299.
12. Shahmanesh M, Gayed S, Ashcroft M, et al.. Geomapping of chlamydia and gonorrhoea in Birmingham. Sex Transm Infect 2000; 76: 268–272.
13. Schleihauf E, Watkins RE, Plant AJ. Heterogeneity in the spatial distribution of bacterial sexually transmitted infections. Sex Trans Infect 2009; 85: 45–49.
14. Gindi RM, Sifakis F, Sherman SG, et al.. The geography of heterosexual partnerships in Baltimore City adults. Sex Transm Dis 2011; 38: 260–266.
15. Sobela F, Pepin J, Gbeleou S, et al.. A tale of two countries: HIV among core groups in Togo. J Acquir Immune Defic Syndr 2009; 51: 216–223.
16. Wylie JL, Jolly A. Patterns of chlamydia and gonorrhea infection in sexual networks in Manitoba, Canada. Sex Transm Dis 2001; 28: 14–24.
17. Zenilman JM, Ellish N, Fresia A, et al.. The geography of sexual partnerships in Baltimore: Applications of core theory dynamics using a geographic information system. Sex Transm Dis 1999; 26: 75–81.
18. Michaud JM, Ellen J, Johnson SM, et al.. Responding to a community outbreak of syphilis by targeting sex partner meeting location: An example of a risk-space intervention. Sex Transm Dis 2003; 30: 533–538.
19. Ogilvie GS, Taylor DL, Moniruzzaman A, et al.. A population-based study of infectious syphilis rediagnosis in British Columbia, 1995–2005. Clin Infect Dis 2009; 48: 1554–1558.
20. 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.
21. Sena AC, Muth SQ, Heffelfinger JD, et al.. Factors and the sociosexual network associated with a syphilis outbreak in rural North Carolina. Sex Transm Dis 2007; 34: 280–287.
22. Jennings JM, Curriero FC, Celentano D, et al.. Geographic identification of high gonorrhea transmission areas in Baltimore, Maryland. Am J Epidemiol 2005; 161: 73–80.
23. Lacey CJ, Merrick DW, Bensley DC, et al.. Analysis of the sociodemography of gonorrhoea in Leeds, 1989–93. BMJ 1997; 314: 1715–1718.
24. Potterat JJ, Rothenberg RB, Woodhouse DE, et al.. Gonorrhea as a social disease. Sex Transm Dis 1985; 12: 25–32.
25. Rice RJ, Roberts PL, Handsfield HH, et al.. Sociodemographic distribution of gonorrhea incidence: Implications for prevention and behavioral research. Am J Public Health 1991; 81: 1252–1258.
26. Sullivan AB, Gesink DC, Brown P, et al.. Are neighborhood sociocultural factors influencing the spatial pattern of gonorrhea in North Carolina? Ann Epidemiol 2011; 21: 245–252.
27. Adimora AA, Schoenbach VJ. Social context, sexual networks, and racial disparities in rates of sexually transmitted infections. J Infect Dis 2005; 191 (suppl 1): S115–S122.
28. Browning CR, Burrington LA, Leventhal T, et al.. Neighborhood structural inequality, collective efficacy, and sexual risk behavior among urban youth. J Health Soc Behav 2008; 49: 269–285.
29. Holtgrave DR, Crosby RA. Social capital, poverty, and income inequality as predictors of gonorrhoea, syphilis, chlamydia and AIDS case rates in the United States. Sex Transm Infect 2003; 79: 62–64.
30. Kilmarx PH, Zaidi AA, Thomas JC, et al.. Sociodemographic factors and the variation in syphilis rates among US counties, 1984 through 1993: An ecological analysis. Am J Public Health 1997; 87: 1937–1943.
31. Semaan S, Sternberg M, Zaidi A, et al.. Social capital and rates of gonorrhea and syphilis in the United States: Spatial regression analyses of state-level associations. Soc Sci Med 2007; 64: 2324–2341.
32. Guillet B, Brown B, Turner CS, et al.. North Carolina Agriculture Regional Perspectives. Raleigh, NC: North Carolina Rural Economic Development Center, 2009:28.
34. Hall SA, Kaufman JS, Ricketts TC. Defining urban and rural areas in U.S. epidemiologic studies. J Urban Health 2006; 83: 162–175.
35. Allshouse WB, Fitch MK, Hampton KH, et al.. Geomasking sensitive health data and privacy protection: An evaluation using an E911 database. Geocarto Int 2010; 25: 443–452.
36. Hampton KH, Fitch MK, Allshouse WB, et al.. Mapping health data: Improved privacy protection with donut method geomasking. Am J Epidemiol 2010; 172: 1062–1069.
37. Kulldorff M. A spatial scan statistic. Commun Stat Theory Methods 1997; 26: 1481–1496.
38. Aamodt G, Samuelsen SO, Skrondal A. A simulation study of three methods for detecting disease clusters. Int J Health Geogr 2006; 5: 15.
39. Song C, Kulldorff M. Power evaluation of disease clustering tests. Int J Health Geogr 2003; 2: 9.
40. Kulldorff M, Athas WF, Feurer EJ, et al.. Evaluating cluster alarms: A space-time scan statistic and brain cancer in Los Alamos, New Mexico. Am J Public Health 1998; 88: 1377–1380.
41. Doherty IA, Adimora AA, Muth SQ, et al.. Comparison of sexual mixing patterns for syphilis in endemic and outbreak settings. Sex Transm Dis 2011; 38: 378–384.
42. Public Health Ontario. Monthly Infectious Diseases Surveillance Report. Toronto: Public Health Ontario; 2012.
43. Torrone EA, Bertolli J, Li J, et al.. Increased HIV and primary and secondary syphilis diagnoses among young men—United States, 2004–2008. J Acquir Immune Defic Syndr 2011; 58: 328–335.
© Copyright 2013 American Sexually Transmitted Diseases Association
44. Richardson D, Goldmeier D. Lymphogranuloma venereum: An emerging cause of proctitis in men who have sex with men. Int J STD AIDS 2007; 18: 11–14; quiz 5.