Gonorrhea is the second most commonly reported nationally notifiable disease, with 301,174 cases reported in the United States in 2009.1 Untreated Neisseria gonorrhoeae infections may lead to serious reproductive health consequences among women including pelvic inflammatory disease, ectopic pregnancy, and infertility. Among men, gonorrhea may result in urethritis and epididymitis.2 Diagnosing, treating, and managing sequelae of gonorrhea in the United States costs up to 212 million dollars annually.3
Gonorrhea trends are largely ascertained from national notifiable disease data. Providers and laboratories are required to report gonorrhea cases to state or local health departments, and these reports are compiled at a national level. Reported case rates have been declining since the mid-1970s. Rates declined 74% from 1975 to 1997, likely because of implementation of a national gonorrhea control program. Case rates then held steady for nearly 10 years, before declining again in 2006. The 2009 rate of 99.1 per 100,000 persons was the lowest since case reporting surveillance began in 1941.1
Although declining case reports may signify a reduction in gonorrhea prevalence, using these data for trend interpretation presents some challenges. Gonorrhea prevalence cannot be directly calculated because only positive test results are reported. In addition, case report data are influenced by changes in reporting practices, screening programs, and test technology. For example, as reporting systems are enhanced, case reports may increase. Screening expansion and more accurate test technology may improve case detection. Conversely, reductions in testing or reporting efforts may decrease the number of gonorrhea case reports. Fluctuations in case report rates caused by such programmatic or structural changes may not reflect true variation in gonorrhea morbidity.
National survey data may be used to supplement case reports and provide more complete information about trends. For example, the National Health and Nutrition Examination Survey indicates that gonorrhea prevalence was 0.24% among the noninstitutionalized US population during the period 1999–2002.4 However, population-based estimates of gonorrhea prevalence are too small and unstable to examine prevalence over time or to assess risk factors for gonorrhea infection. In order to improve our understanding of gonorrhea trends over time, screening data from the National Job Training Program (NJTP) were used to estimate gonorrhea prevalence in a high-risk, stable population.
MATERIALS AND METHODS
The NJTP provides career preparation training for economically disadvantaged 16- to 24-year-old men and women. Since its establishment under the US Department of Labor in 1964, more than 2 million young people have received training in 48 states, the District of Columbia, and Puerto Rico.5 In program year 2008 (July 2008–June 2009), 60,896 men and women were enrolled in 124 NJTP residential training centers. Of these participants, 59% were male, and 80% were <21 years of age. Most were racial/ethnic minorities; 52% were black, and 17% were Hispanic. Only 28% had a high school degree or equivalent upon entry to the program.6
The NJTP requires screening for a number of health conditions, including gonorrhea and chlamydia, within 2 days of entry.7 Men were tested for gonorrhea and chlamydia using the same nucleic acid amplification test (NAAT), the strand displacement amplification test (SDA; BD ProbeTec ET, Becton-Dickinson, Sparks, MD), on first void urine specimens throughout the study period. Women are also tested using SDA technology on either cervical or urine specimens. However, before 2006, women were tested using a DNA hybridization probe (DNA probe, PACE 2, Gen-Probe, San Diego, CA) on cervical specimens and SDA on urine specimens.
One federally contracted laboratory (Lab A) tested all specimens for chlamydia and the majority of specimens for gonorrhea.8 All data used in this analysis were generated from Lab A. In the minority of NJTP sites where Lab A was not used for gonorrhea testing, sites either contracted with other laboratories to perform gonorrhea testing or did not universally screen program entrants, despite NJTP recommendations.
To approximate gonorrhea screening and minimize the influence of centers that tested only symptomatic or high-risk entrants for gonorrhea, the potential bias resulting from diagnostic gonorrhea testing was examined. Sex-specific gonorrhea prevalence was compared between all NJTP entrants and entrants into centers testing minimum percentages of entrants for both chlamydia and gonorrhea—50%, 75%, and 90%. On the basis of results from this analysis, which suggested similar gonorrhea prevalence in all 3 groups, the analytic sample was limited to training centers testing at least 50% of entrants for both chlamydia and gonorrhea.
Gonorrhea prevalence trends were estimated among men and women entering the NJTP, between 2004 and 2009, using laboratory data from 48 states and the District of Columbia. These data included basic demographics (sex, age, race/ethnicity, region), gonorrhea test technology used, and test result.
Gonorrhea prevalence was calculated by dividing the number of people testing positive for gonorrhea by the total number of people tested in the analytic sample. First, prevalence was estimated by sex and year. To assess how gonorrhea prevalence varied demographically, sex-specific annual prevalence was also calculated by age group, race, region, and test type (for women). χ2 tests for independence were used to assess statistical differences by year and all other variables.
Multivariate logistic regression was used to assess trends in gonorrhea prevalence over time, modeling gonorrhea test result as the dependent variable and year as the independent variable of interest. Linearity of the relationship between year and gonorrhea test result was assessed using logit plots, which indicated year could be modeled continuously. Models were estimated separately for men and women and adjusted for variables previously demonstrated to be associated with gonorrhea risk. Variables for models of both men and women included age, race/ethnicity, and region, and the women's model also included test type. Condition indices were evaluated to assess collinearity between independent variables. Interactions between year and all other independent variables were tested. The Hosmer-Lemeshow goodness-of-fit test was used to check model fit.
Sensitivity analyses were performed to examine whether annual gonorrhea prevalence estimates were affected by inclusion of different training centers over time. Annual prevalence estimates were calculated separately for men and women entering only centers that screened 50%, 75%, and 90% of entrants in all 6 years. Logistic regression was performed using each of these samples separately, and odds ratios of testing positive for gonorrhea by year were compared across models.
Between 2004 and 2009, 212,117 men and 135,328 women entered 124 NJTP residential training centers. Of these, 113,609 men (53.6% of total) and 105,964 women (78.3% of total) were tested for gonorrhea upon entry to the program. After excluding 11,968 men (10.5%) and 11,588 women (10.9%) missing data on race or gonorrhea test result, and further excluding 6457 men (5.7%) and 2679 (2.5%) women entering centers not uniformly screening for gonorrhea (screening <50% of those screened for chlamydia), the analytic sample included 95,184 men and 91,697 women. This sample represents 83.8% of all men tested for gonorrhea and 86.6% of all women tested. The sample increased each year, ranging from 9977 men and 11,728 women in 2004 to 18,882 men and 17,910 women in 2009.
Table 1 shows characteristics of men and women screened for gonorrhea between 2004 and 2009. More than 70% of men and women were <20 years of age. More women than men were racial/ethnic minorities; 77% of women and 72% of men were black or Hispanic. The population was geographically concentrated in the South, where nearly 56% of men and 48% of women resided.
Figure 1 shows unadjusted gonorrhea prevalence by sex and year. Women had higher gonorrhea prevalence than men in each year from 2004 to 2009. Among women, gonorrhea prevalence increased from 2.6% in 2004 to 2.9% in 2006, then decreased steadily to 1.8% in 2009. For men, prevalence was 1.3% in 2004, peaking at 1.6% in 2005 then declining to 0.9% in 2009.
Tables 2 and 3 show gonorrhea prevalence stratified by age, race/ethnicity, region, and test type (for women). Among men, gonorrhea declined significantly over time among those aged 16 to 21 years, whites, blacks, and those residing in the South. Gonorrhea prevalence was markedly higher among black men (2.0%) than among Hispanic men (0.5%) or white men (0.2%) in 2004. Though this racial disparity persisted in 2009, prevalence among black men decreased to 1.5%. Geographically, gonorrhea prevalence decreased most in the highest prevalence areas—from 2.0% to 1.1% in the Midwest and from 1.5% to 1.1% in the South—between 2004 and 2009. Gonorrhea either decreased or remained stable over time in every age, racial/ethnic, or regional subgroup.
Among women, gonorrhea prevalence declined significantly from 2004 to 2009 among those aged 18 to 21 years, whites, blacks, and those residing in the South and West. Prevalence was higher among women in the 16- to 17-year-old age group than among those in other age groups, nearly every year. Prevalence in this age group decreased, though nonsignificantly, from 3.1% to 2.4% between 2004 and 2009.
Between 2004 and 2008, black women had more than 3 times the gonorrhea prevalence of white women; however, prevalence among black women decreased markedly from 3.6% in 2004 to 2.5% in 2009. Like men, women residing in the South and Midwest were more likely to test positive for gonorrhea than those residing in other areas. Prevalence declined from 3.4% to 2.6% among women in the Midwest and from 3.1% to 2.2% among women in the South.
Table 3 also shows that from 2004 to 2005 most women were tested for gonorrhea using the cervical DNA hybridization probe. In 2006, the NJTP transitioned to using the NAAT on either urine or cervical specimens, and by 2007, very few women were tested using the DNA hybridization probe. Between 2005 and 2009, prevalence was highest among women tested using the NAAT on cervical specimens, though gonorrhea prevalence decreased from 8.9% to 2.2% of women tested using this technology. The high prevalence observed in 2005 likely reflects early, potentially diagnostic, use of cervical NAATs for women at high risk for gonorrhea infection.
Table 4 shows results from unadjusted and adjusted logistic regression assessing the association between gonorrhea prevalence and year. Significant linear trends for decreased gonorrhea prevalence were observed over time for men and women. After adjusting for age, race, and region, the odds of men testing positive for gonorrhea decreased by 8% each year, translating to a 40% decrease in odds over the 5-year study period. After adjusting for age, race, region, and test technology, the odds of women testing positive for gonorrhea decreased by 10% for each year of study, or by 50% from 2004 to 2009. Interaction terms between year and other variables were not significant in models of either men or women. Final models adequately fit the data, demonstrated by Hosmer-Lemeshow goodness-of-fit χ2 tests, P = 0.24 for the men's model and P = 0.93 for the women's model.
The sensitivity analyses including only centers continuously screening 50%, 75%, and 90% of entrants between 2004 and 2009 revealed comparable annual gonorrhea prevalence estimates to those reported here. Three additional multivariate logistic regression models were estimated using each of these continuously screening samples. Each yielded similar and statistically significant reductions in the odds of testing positive for gonorrhea by year.
Gonorrhea prevalence declined from 2004 to 2009 among young men and women entering the NJTP. After controlling for changes in entrants' demographic characteristics, a 40% decrease was seen among men, compared to a 50% decrease among women. To our knowledge, this is the first analysis examining trends over time in gonorrhea prevalence among NJTP entrants. By allowing prevalence estimation across multiple years in a large, stable population, these data present a unique opportunity to better understand the changing gonorrhea epidemic.
The decrease in gonorrhea prevalence among NJTP entrants suggests that the observed decline in national case reports over the same period were at least partially due to a real decrease in population prevalence. From 2004 to 2009, case reports declined substantially. More than 112 cases per 100,000 persons were reported in the US population in 2004, compared with 99 cases per 100,000 persons in 2009. During this time period, case rates peaked in 2006, at nearly 120/100,000 persons.1 Case rates also declined over this time period among young adults aged 15 to 24 years, declining from 455/100,000 to 441/100,000 persons.
The use of dual NAAT technology, which simultaneously tests for chlamydia and gonorrhea, likely impacted gonorrhea trends in both case report and NJTP data. The NAAT technology may affect gonorrhea trends in 2 ways: through better detection of infections and increased screening. In the NJTP, laboratorians started using NAAT technology in 2006, around the same time as many other laboratorians throughout the country.9–11 Because NAATs detect both chlamydia and gonorrhea with higher sensitivity than previous test technologies, gonorrhea prevalence in the NJTP may have artificially increased in 2006.
Use of dual NAATs may have also impacted gonorrhea prevalence in the general population and may be one explanation for declining trends. CDC has recommended annual chlamydia screening of sexually active young women since 1993.12 As the dual NAAT came into widespread use, women already being screened for chlamydia were also tested for gonorrhea. Case rates likely increased in 2006 as a result of more screening. However, this increase may have been followed by a real decline in population-level gonorrhea prevalence because women with asymptomatic, prevalent infections were diagnosed and treated.
Declining annual trends among black men and women were particularly encouraging, given the disproportionately high burden of disease in these groups. Still, rates were 5 times higher in 2009 among black men compared with white men and nearly 3 times higher among black women compared with white women. This racial disparity has been demonstrated in numerous other studies, as well as through national surveillance. Reported gonorrhea case rates were 20 times higher among blacks than whites in 2009.1 Because case report data are partially driven by who is screened for gonorrhea, prevalence data may provide a more accurate estimate of racial disparities in gonorrhea risk. More consistent with the NJTP findings, the National Health and Nutrition Examination Survey data indicate that the national prevalence of gonorrhea was approximately 4 times higher among blacks compared with whites between 1999 and 2006.13 Likewise, among women screened in family planning clinics, gonorrhea positivity was 6 times higher among blacks compared with whites.14
The highest gonorrhea prevalence observed in our study was among young black women. National case reporting data also indicate that black women aged 15 to 24 have the highest risk of any racial or age group for both gonorrhea and chlamydia.1 The decline in gonorrhea prevalence observed in this group is encouraging, but targeted prevention interventions are urgently needed to address this persistently disproportionate burden of disease.
In addition to race, previous research suggests gonorrhea infection is also associated with young age 4,15–17 and residing in the South.16 By virtue of their demographic characteristics, NJTP entrants are at higher risk for gonorrhea infection than the general US population. Entrants are young and economically disadvantaged, as well as disproportionately black and likely to reside in the southern region of the United States.
The NJTP data are uniquely suited to examine gonorrhea trends over time. Because NJTP entrants have higher gonorrhea risk than the general population as captured in national survey samples, prevalence can be stably estimated within regional and racial subgroups, as well as over multiple years. Although this is the first study to describe gonorrhea prevalence time trends among NJTP entrants, a previous study examined gonorrhea prevalence and risk factors among young adults entering the NJTP in 1996.18 Studies using NJTP data to examine trends in chlamydia prevalence have similarly found declines over time.19,20 NJTP studies have also found chlamydia was associated with black race and young age.20–22
This analysis has several limitations. First, information on behavioral risk factors and sexual orientation were not available as part of this data set, making it impossible to examine how associations between demographic characteristics and gonorrhea infection may be mediated by behavior. Second, it is possible that diagnostic gonorrhea tests given to symptomatic individuals were included in the analysis. By limiting the analysis to centers with data on at least 50% of entrants for both chlamydia and gonorrhea, gonorrhea screening was approximated. Although sensitivity analyses indicated similar gonorrhea prevalence using this 50% cut point and a 90% cut point, it remains uncertain whether every test was administered for screening purposes. Last, NJTP entrants are not representative of all US young adults and are at higher risk for gonorrhea on the basis of their demographic characteristics.
Despite these limitations, this analysis sheds light on the changing gonorrhea epidemic. The NJTP population is relatively stable, and the available data facilitate adjustment for demographic characteristics that changed slightly over time. The same laboratory performed all gonorrhea tests and provided data for all 6 years, thus minimizing diagnostic and information bias. It is important to note that the NJTP data provide one of the few sources from which gonorrhea prevalence can be estimated. These data provide a comparison to case reporting data, albeit indirect, and allow for triangulation of national surveillance trends.
These data suggest that gonorrhea screening guidelines need refining, with a particular focus on helping providers better understand the local epidemiology of gonorrhea in their practice area in terms of geographic and racial disparities. CDC does not recommend universal gonorrhea screening as is recommended for chlamydia, and the relatively low prevalence of gonorrhea even in high risk subgroups of the NJTP population suggests universal screening is not warranted. However, the US Preventative Services Task Force (USPSTF) recommends that sexually active women should be screened for gonorrhea if they are at “increased risk for infection,” citing young age, black race, previous infection, new sexual partners, inconsistent condom use, and high prevalence of gonorrhea in the patient's community of residence among potential risk factors.23 Physicians are unlikely to have enough information to make decisions about screening on the basis of this recommendation.
These data suggest that the NJTP is an effective screening venue for diagnosing and treating gonorrhea among a high-risk population. Despite decreasing gonorrhea prevalence among NJTP entrants from 2004 to 2009, and potentially in the US population over the same time period, some groups remain at much higher risk than others. Population-level interventions are needed to reduce gonorrhea infections among adolescents and young adults. In particular, interventions should be targeted to reach populations with disproportionate risk.
The extent to which declining gonorrhea case reports signify a real reduction in gonorrhea morbidity, versus decreased testing or reporting, is unknown. However, this analysis suggests that recent declines in gonorrhea case counts reflect declining population prevalence. Using discrete data sources that allow calculation of gonorrhea prevalence is essential for validating case report trends and improving surveillance. Data sources that facilitate examination of gonorrhea risk by demographic subgroups are particularly valuable for focusing prevention and treatment resources on those who need them most.
1.Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance 2009. Atlanta,GA: U.S. Department of Health and Human Services, 2010.
2.Holmes KK, Sparling P, Stamm W, et al. Sexually Transmitted Disease. 4th ed. New York, NY: McGraw-Hill, 2008.
3.Pultorak E, Wong W, Rabins C, et al. Economic burden of sexually transmitted infections: Incidence and direct medical cost of chlamydia, gonorrhea, and syphilis among Illinois adolescents and young adults, 2005–2006. Sex Transm Dis 2009; 36:629–636.
4.Datta SD, Sternberg M, Johnson RE, et al. Gonorrhea and Chlamydia in the United States among persons 14–39 years of age, 1999 to 2002. Ann Intern Med 2007; 147:89–95.
7.US Department of Labor, Office of Job Corps. Policy and Requirements Handbook. Washington, DC: US Department of Labor, November 23, 2010. Available at www.jobcorps.gov/Libraries/pdf/prh.sflb
. Accessed November 30, 2010.
9.Dicker LW, Mosure DJ, Steece R, et al. Laboratory tests used in US public health laboratories for sexually transmitted diseases, 2000. Sex Transm Dis 2004; 31:259–264.
10.Dicker LW, Mosure DJ, Steece R, et al. Testing for sexually transmitted diseases in U.S. public health laboratories in 2004. Sex Transm Dis 2007; 34:41–46.
11.Centers for Disease Control and Prevention. Volume and Type of Laboratory Testing Methods for Sexually Transmitted Disease in Public Health Laboratories, 2007. Atlanta, GA: U.S. Department of Health and Human Services, 2011.
12.Centers for Disease Control and Prevention. Recommendations for the prevention and management of Chlamydia trachomatis
infections, 1993. MMWR 1993; 42(RR-12).
13.Johnson RE, Tian LH, Datta SD, et al. Prevalence of Neisseria gonorrhoeae
infection in the United States non-institutionalized population ages 14–39, 1999–2006. Presented at: International Society for STD Research; 2009; London, United Kingdom.
14.Gorgos L, Newman L, Satterwhite C, et al. Gonorrhoea positivity among women aged 15–24 years in the USA, 2005–2007. Sex Transm Infect 2011; 87:202–204.
15.Mertz KJ, Schwebke JR, Gaydos CA, et al. Screening women in jails for chlamydial and gonococcal infection using urine tests. Sex Transm Dis 2002; 29:271–276.
16.Dicker LW, Mosure DJ, Berman SM, et al. Gonorrhea prevalence and coinfection with chlamydia in women in the United States, 2000. Sex Transm Dis 2003; 30:472–476.
17.Geisler WM, James AB. Chlamydial and gonococcal infections in women seeking pregnancy testing at family-planning clinics. Am J Obstet Gynecol 2008; 198:502e1–502e4.
18.Lifson AR, Halcon LL, Hannan P, et al. Screening for sexually transmitted infections among economically disadvantaged youth in a National Job Training Program. J Adolesc Health 2001; 28:190–196.
19.Mertz KJ, Ransom RL, St Louis ME, et al. Prevalence of genital chlamydial infection in young women entering a National Job Training Program, 1990–1997. Am J Public Health 2001; 91:1287–1290.
20.Satterwhite CL, Tian LH, Braxton J, et al. Chlamydia prevalence among women and men entering the National Job Training Program: United States, 2003–2007. Sex Transm Dis 2010; 37:63–67.
21.Joesef MR, Mosure DJ. Prevalence trends in chlamydial infections among young women entering the National Job Training Program, 1998–2004. Sex Transm Dis 2006; 33:571–575.
22.Satterwhite CL, Joesoef MR, Datta D, et al. Estimates of Chlamydia trachomatis
infections among men: United States. Sex Transm Dis 2008; 35:S3–S7.
23.U.S. Preventive Services Task Force. Screening for gonorrhea: Recommendation statement. Ann Fam Med 2005; 3:263–267.