Introduction

According to the Centers for Disease Control and Prevention (CDC), gonorrhea is the second most commonly reported identifiable disease in the United States (U.S.) [1, 2]. At the same time, the CDC estimates that more than 25% of new Neisseria gonorrhoeae infections demonstrate antibiotic resistance and has classified gonorrhea as an “urgent threat” [3]. Importantly, inadequately treated infections often result in serious complications and long-term morbidity, while infected individuals may remain asymptomatic for extensive periods thus facilitating disease transmission [4]. Moreover, gonococcal infections promote the transmission of HIV thus exerting a disproportionately severe, indirect impact on the reproductive health of the community [5, 6].

Given the scarcity of available public health resources, effective interventions must be targeted at high-risk populations. Such a strategy would be particularly effective against gonorrhea, as wide variations in disease incidence exist in different sociodemographic groups [7, 8]. Even though several studies have examined the influence of sociodemographic variables on the incidence and outcome of gonorrhea [4, 7, 9–13], little information exists on the precise impact and relative importance of these factors and the association of gonorrhea infection rates with the incidence of HIV on a national scale. In this cross-sectional study, we utilized nationwide data from the National Center for HIV/AIDS, Viral Hepatitis, Sexually Transmitted Diseases, and Tuberculosis Prevention (NCHHSTP) Atlas database [14] and the U.S. Census Bureau [15], in order to examine the association of gonorrhea with socioeconomic variables and HIV incidence. We then constructed a multivariable model of gonorrhea incidence on the basis of the identified variables.

Methods

Results and discussion

A total of 1,638,863 N. gonorrhoeae infections were reported in 3,220 counties or county-equivalents (from the U.S., Puerto Rico, the U.S. Virgin Islands and Guam) between 2010 and 2014. The adjusted average gonorrhea incidence rate was 103.30 cases per 100,000 people (95% CI: 103.14, 103.45) and the incidence of gonorrhea increased from 100.19 cases per 100,000 people in 2010 to 109.79 cases per 100,000 people in 2014. The District of Columbia [303.26 cases per 100,000 people (95% CI: 297.21, 309.40)], Mississippi [197.17 cases per 100,000 people (95% CI: 194.92, 199.43)] and Louisiana [194.10 cases per 100,000 people (95% CI: 192.31, 195.91)] had the highest reported rates, while New Hampshire [11.74 cases per 100,000 people (95% CI: 10.92, 12.59)], Wyoming [10.86 cases per 100,000 people (95% CI: 9.69, 12.13)] and Puerto Rico [9.90 cases per 100,000 people (95% CI: 9.45, 10.37)] had the lowest rates of N. gonorrhoeae infection. Overall, the highest incidence rates were concentrated in the Southern states (Figs 1 and 2, Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. N. gonorrhoeae infection rates across the United States.

Choropleth Map of Gonorrhea Incidence in the United States, 2010–2014.

https://doi.org/10.1371/journal.pone.0183938.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. N. gonorrhoeae infection hotspots in the United States.

Gonorrhea Incidence Hotspot Analysis in the United States, 2010–2014.

https://doi.org/10.1371/journal.pone.0183938.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. N. gonorrhoeae infection rates by state, 2010–2014.

https://doi.org/10.1371/journal.pone.0183938.t001

Repeated simple linear regression was performed to examine the association of a number of independent variables with gonorrhea incidence rates. Average HIV incidence rates, percentage of women, percentage of individuals of Black race, percentage of individuals under the poverty line and county urban ranking were positively associated with the incidence of gonorrhea (Table 2, all variables were population-adjusted).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Univariable and multivariable linear regression analysis of N. gonorrhoeae infection rates at the county level.

https://doi.org/10.1371/journal.pone.0183938.t002

Subsequently, the independent variables identified through the univariable analysis were combined to produce a multivariable linear regression model. HIV incidence rates, county urban ranking, percentage of women and percentage of individuals under the poverty line (Table 2) were independent predictors of gonorrhea incidence (Pmodel<0.001, R² model = 0.56).

Notably, although Black race has previously been associated with increased risk of gonorrheal infection [8, 10, 11, 13, 20], we elected not to include race in the final model. This decision stemmed from the fact that race primarily reflects disparities in access to healthcare and the overall standard of living, as evidenced by the collinearity that race displayed with other demographic variables. In turn, this finding suggests that race could often simply function as a proxy of socioeconomic conditions, rather than a true independent risk factor and should be used with caution in epidemiologic studies [21, 22]. A patient-level analysis may be better suited to identifying the relative impact of the socioeconomic and biologic risk factors (such as absence of the CCR5 mutation in individuals of black race, leading to increased susceptibility to HIV [23–25]) that might underlie the association of race with the incidence of STIs.

Importantly, HIV incidence rates had a powerful association with N. gonorrhoeae infection. Sexually transmitted infections not only share common risk factors with HIV infection, but also facilitate HIV transmission [5, 26] and previous studies have reported that N. gonorrhoeae infection may result in up to a 3-fold increase in the risk of HIV seroconversion [6, 27, 28]. This is noteworthy, given the morbidity and healthcare costs associated with HIV infection. For example, a recent study estimated the lifetime cost of HIV infection at approximately $326,500 [29] and, as these figures were based on conservative assumptions, overall costs of HIV infection may be even higher in practice. Although previous trials that assessed the impact of treatment for STIs on HIV incidence produced inconclusive results [30, 31], some evidence suggests that timely therapy for STIs, including gonorrhea, may help to decrease HIV transmission, particularly in the setting of high STI and HIV incidence [32, 33].

Moreover, our findings emphasize the importance of preventive measures whenever an STI, such as gonorrhea, is diagnosed. For example, given the proven efficacy and relative safety of pre-exposure prophylaxis (PrEP) for HIV, health care providers should consider whether PrEP should be offered to patients with a history of an STI [34–36]. Recent recommendations of the International Antiviral Society-USA Panel suggest that PrEP should be discussed with patients that were recently diagnosed with an STI, particularly if their yearly risk of contracting HIV is 2% or more [34, 36]. Interestingly, a recent study by Solomon et al. suggested that Men who have Sex with Men and are diagnosed with syphilis are at very high risk for HIV seroconversion and may constitute ideal candidates for PrEP [37]. Although it may be tempting to hypothesize that an association may also exist between N. gonorrhoeae infection and subsequent HIV seroconversion, the present population-based study cannot establish causation or determine the nature and direction of the association between gonorrhea and HIV infection on an individual level. Patient-level analyses are needed to investigate the interplay of HIV and N. gonorrhoeae infection and determine the precise indications of PrEP administration in this patient group. Moreover, it must be noted that the decision to initiate PrEP does not obviate the need for patient education and appropriate counseling on sexual behavior, especially given the minimal cost and risk-free nature of such interventions [38].

This multivariable analysis also demonstrated that county poverty levels were independently associated with gonorrhea incidence rates. Socioeconomic disadvantage has been shown to exert an impact on N. gonorrhoeae infection [8, 11], and our results lend further credence to previous studies, by utilizing data on a national scale. Although the present analysis cannot conclusively demonstrate whether the reported association is directly causal, income status may serve as a proxy of community sexual health. Poverty is inextricably associated with low educational attainment [39], decreased access to health care [40], substance abuse [41, 42] and increased prevalence of prostitution [43], all of which are known to either promote high-risk sexual behavior or prevent the early identification and treatment of N. gonorrhoeae infection [10, 13]. Although some of these factors may appear challenging to address, positive changes in socioeconomic conditions may substantially impact gonorrhea incidence rates [11].

Lastly, our findings also point to increased gonorrhea incidence in urban counties, as well as in counties with a greater percentage of women. The first finding is consistent with the existing literature, which suggests that gonorrhea tends to be more frequent in the urban setting, particularly in areas characterized by increased population density and considerable socioeconomic disadvantage [10, 44]. Nevertheless, we must note that gonorrhea infection rates may be consistently underreported in rural areas [12]. As such, the importance of public health initiatives in the rural setting could be underestimated. Regarding the higher overall incidence of gonorrhea in counties with a larger female population, this finding may partly be attributed to the fact that women commonly develop asymptomatic infection [4]. In turn, although the rate of asymptomatic infection is not adequately captured by incidence statistics, asymptomatic individuals are central to disease transmission as they may unwittingly serve as a disease “reservoir”. In fact, mathematical models suggest that the presence of a preexisting high-incidence “reservoir” of N. gonorrhea infection in a sexual network may have a greater impact on disease spread than individual sexual behavior [45]. As such, our findings underlie the importance of targeted screening programs in high-risk women, as outlined in the recommendations recently issued by the U.S. Preventive Services Taskforce [46].

Regarding limitations that should be considered before interpreting our findings, the present analysis relied on aggregate incidence rates and sociodemographic data, collected at the county and state level, rather than data from individual cases. As such, detailed patient-level stratification according to possible confounders was impossible to perform. Furthermore, the cross-sectional design of the study precluded the determination of cause-and-effect relationships, but consideration of the existing literature facilitates interpretation of our findings. The databases we utilized enabled us to perform a nationwide analysis and limited the possibility of sampling bias. However, the timeframe of our analysis, as well as the availability of pertinent sociodemographic variables, were limited. Moreover, the possible existence of a systematic pattern of underreporting among counties with specific sociodemographic characteristics may have introduced bias into our analysis.

Conclusions

In conclusion, the incidence of gonorrhea in the U.S. increased over the 2010–2014 time period. Given the high frequency of asymptomatic infection [4], the increasing rate of antibiotic resistance among N. gonorrhoeae strains [2] and the association of gonorrhea with HIV transmission [5], a comprehensive public health response is necessary to contain disease spread. We developed a county-level, multivariable linear regression model of N. gonorrhoeae infection, after identifying independent predictors of gonorrhea incidence on a national scale. Importantly, our analysis demonstrated that gonorrhea and HIV infection are closely associated at the population level. Although the present study cannot conclusively prove whether this association is also valid at the patient level, it suggests the potential benefits from comprehensive screening for STIs and HIV. This is particularly important, given the large number of unreported/undiagnosed STI cases [3, 47] and the low compliance with screening recommendations that is reported in the literature [48–51]. Furthermore, our findings highlight the potential value of offering pre-exposure prophylaxis for HIV to high-risk patients visiting an STI clinic, particularly in high-incidence areas. Sociodemographic factors were also associated with gonorrhea incidence, thus suggesting another possible focus for public health initiatives. Taken together, a combination of targeted screening programs, counseling on sexual behavior and comprehensive clinical management of high-risk patients with STIs may prove instrumental in curtailing the spread of gonorrhea and HIV infection. Future studies that focus on the most heavily affected areas will provide additional guidance on how to allocate scarce public health funds and design cost-effective initiatives.