PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of pubhealthrepPublic Health Reports
 
Public Health Rep. 2009; 124(Suppl 2): 18–23.
PMCID: PMC2775396

Trends in Neisseria gonorrhoeae Incidence Among HIV-Negative and HIV-Positive Men in Washington State, 1996–2007

SYNOPSIS

Objectives

We assessed population-level trends in human immunodeficiency virus (HIV) and Neisseria gonorrhoeae co-infection among adult males in Washington State between 1996 and 2007.

Methods

Population-based categorical disease surveillance registries for gonorrhea and for HIV were electronically matched and merged at the record level and incidence rates were calculated for reported HIV-positive and presumed HIV-negative men.

Results

The incidence of gonorrhea infection increased significantly among both HIV-positive and presumed HIV-negative men from 1996 to 2005, and this trend has recently reversed for both groups. The annual incidence rate of gonorrhea among HIV-positive men was found to be higher in all years than among men presumed to be HIV-negative.

Conclusions

Inequality in the burden of gonorrhea by HIV-infection status suggests continuing sexual risk-taking among HIV-positive men as well as possible barriers to diagnosis, treatment, and partner services. This inequality may also reflect significant differences in gonorrhea burden among men who have sex with men as well as by HIV status. Monitoring emergent secular trends in population-level HIV/sexually transmitted infection comorbidity through registry matching is an achievable strategy for developing an evidence base to inform program collaboration and service integration efforts aimed at providing more comprehensive disease prevention services.

Considerable evidence suggests that co-infection with human immunodeficiency virus (HIV) and bacterial sexually transmitted diseases (STDs) can act synergistically to facilitate HIV transmission.1 Clinical studies have demonstrated that infection with Neisseria gonorrhoeae (N. gonorrhoeae) can significantly amplify the concentration of HIV in seminal fluid2 and can promote HIV replication in human dendritic cells.3 Dendritic cells are likewise known to serve as a reservoir for delivery of virus to activated CD4+ T lymphocytes.4 These observations suggest that N. gonorrhoeae may play an epidemiologically meaningful role in HIV transmission at the population level. Moreover, incident N. gonorrhoeae co-infection in an individual with HIV disease provides an unambiguous biological marker for recent sexual risk behavior, which is of interest to HIV prevention efforts at both the individual and aggregate levels.

The Centers for Disease Control and Prevention (CDC) identifies prevention for HIV-positive individuals as a key strategy for preventing new domestic HIV infections.5 Both HIV and gonorrhea are nationally notifiable diseases, and routine, nonidentified public health surveillance data are readily available. These data are used to describe overall trends in HIV infection and gonorrhea incidence. However, case report data are not integrated by individual patient in national registries, severely limiting the utility of these public health surveillance data for assessing trends in co-infection.

Systemic, technological, and programmatic barriers exist to integrating case surveillance data at the state level, many of which have historical antecedents. These barriers constrain the integration of surveillance data in many jurisdictions. Additionally, structural differences in surveillance systems and coding schema present a challenge to integrating records at the patient level. Many health departments use a CDC-developed database called STD*MIS to manage STD surveillance information.6 HIV/acquired immunodeficiency virus (AIDS) surveillance data are managed using the CDC-mandated HIV/AIDS Reporting System (eHARS).7 Differences in file structures between these two systems necessitate external processing to match and merge records. Differences in how patient information is collected and coded also present challenges. For example, until relatively recently, the schema for coding race and ethnicity data differed significantly between STD and HIV surveillance systems, making these data difficult to use as criteria in matching records.

Differences, both real and perceived, in data security and confidentiality standards between HIV and STD programs at the state level also present a barrier to assessing the extent of comorbidity based on merged surveillance records. This is the situation in Washington State as elsewhere in the U.S., where provisions in state administrative codes strictly govern the uses of HIV surveillance data while similar provisions do not exist for STD-related data. This article describes methods used in Washington State to integrate HIV/STD registries in light of these concerns and describes findings from our analysis of HIV/gonorrhea co-infection.

METHODS

We obtained records for cases of HIV infection from HARS for male case subjects diagnosed through December 31, 2007, and reported as of July 31, 2008. All male case subjects aged 18 years and older, based on current age by year, were included regardless of whether they were initially diagnosed in Washington State or reported from other jurisdictions. A total of 18,627 cases of HIV or AIDS met these inclusion criteria. Annual denominators for rate calculations included HIV-positive males, aged ≥ 18 years, and not known to be deceased for each year of the study period. Gonorrhea morbidity records were obtained from STD*MIS for 18,313 cases among males aged ≥ 18 years at diagnosis and diagnosed between January 1, 1996, and December 31, 2007.

To address confidentiality concerns in Washington State, we chose to integrate HIV and STD surveillance data for this analysis with no permanent link created in either categorical disease registry. Personal identifiers for matched records were not retained in the merged dataset, preserving the historical integrity of unitary and separate registries for HIV and STD surveillance.

To construct a temporary dataset of matched records, we used a scored, deterministic matching algorithm based on the first three letters of the patient's last name; first two letters of the first name; and month, day, and year of birth, according to the scoring schema presented in Figure 1.

Figure 1
Registry matching criteria and scoring used to integrate STD and HIV surveillance data, Washington State, 1996–2007

By this scoring schema, a total of 90 points were available to indicate the strength of the match between any two record pairs. Based on manual review of the match output, we determined that all matching pairs with a point value of 80 or higher indicated true matching records. Pairs with point values lower than 50 were found to contain a large proportion of false matches and were considered nonmatching for the purposes of this study. Records scoring between 50 and 80 points were manually reviewed and coded as a match or nonmatch based on manual inspection of the record pairs.

Matched patient records were merged into a de-identified dataset for analysis that included HIV and STD diagnoses dates and characteristics of reported STD diagnoses. We constructed a binary variable indicating the HIV status of the patient (presumed HIV-negative vs. known HIV-positive) at the time of the STD diagnosis, based on the earliest indication of HIV infection in the HIV surveillance record. We use the term “presumed HIV-negative” to indicate men not diagnosed and reported to Washington State's HIV surveillance registry; true HIV status of these men is not known, but no surveillance evidence exists to indicate that they were infected at the time of their gonorrhea diagnosis.

We constructed a second dataset containing records for all reported HIV-positive males, including HIV exposure risk, patient demographics, and variables calculated from the registry match indicating presence of a gonorrhea diagnosis and whether the diagnosis was subsequent to their earliest indication of HIV infection or prior to their HIV diagnosis. All matches, statistical calculations, and analysis datasets were produced using SAS® version 9.1.2.8

Using these matched datasets, we calculated annual gonorrhea incidence rates per 100,000 among HIV-positive men aged ≥ 18 years. Gonorrhea cases diagnosed annually among the surveillance cohort of reported HIV-positive men provided the numerator for our rate calculation, while the number of men aged ≥ 18 living with HIV in each calendar year, based on cases reported to the HIV case registry, served as the denominator. We obtained age, race, and risk characteristics for HIV-positive gonorrhea cases from the HIV case report.

Exact Poisson confidence intervals were calculated for all rates. Population denominators for presumed HIV-negative men were calculated by subtracting the number of men known to be HIV-positive each year from the total adult male population estimates developed based on U.S. Census counts for Washington State. We used ordinary least-squares regression to assess significance of trends in annual incidence among HIV-positive and presumed HIV-negative groups. The annual percent change and absolute rate difference between HIV-positive and presumed HIV-negative groups was calculated, plotted, and evaluated for magnitude, direction, and significance of observed trends. We assessed single-year changes in incidence and differences in demographics using the Chi-square test of significance.

RESULTS

We identified 1,469 cases of gonorrhea between 1996 and 2007 among men aged ≥ 18 years known to be HIV-positive at the time of their gonorrhea diagnosis, and 16,844 cases among men presumed to be HIV-negative across the entire study period. Demographic and risk characteristics of men with gonorrhea cases identified subsequent to HIV infection are presented in the Table. HIV-risk exposure categories included male-to-male sexual contact (men who have sex with men [MSM]), injection drug use (IDU), heterosexual contact (HETERO), no reported risk (NIR), and other risk such as contaminated blood products. The overwhelming majority of these cases reported MSM or MSM/IDU risk, accounting for 94.9% of incident gonorrhea cases among HIV-positive men across the study period. We also found marked demographic differences in gonorrhea incidence by HIV status across the study period. HIV-positive men with gonorrhea were significantly more likely to be white after adjusting for missing race on STD case reports (odds ratio [OR] estimate = 2.53, range: 2.24–2.87, p<0.00001) and more likely to be aged ≥ 30 years (OR estimate = 4.53, range: 3.96–5.19, p<0.00001) compared with presumed HIV-negative men with gonorrhea.

Table
Demographic and HIV risk characteristics of incident gonorrhea cases among men aged ≥ 18 years by HIV status, Washington State, 1996–2007

Gonorrhea incidence among both presumed HIV-negative (Figure 2) and HIV-positive (Figure 3) males increased significantly from 1996 through 2005. This trend peaked for HIV-positive males in 2005, and by 2007 the incidence rate had returned to a level not significantly higher than the level initially observed in 1996. Among presumed HIV-negative males, gonorrhea incidence reached a zenith of 97.6 cases per 100,000 men in 2006 before decreasing significantly to 71.2 cases per 100,000 men in 2007. Despite this apparent and welcome downward inflection in gonorrhea incidence trends for both groups, the rate of gonorrhea among HIV-positive men was observed to be consistently and considerably higher than among presumed HIV-negative men across the entire study period.

Figure 2
Gonorrhea rate and incidence trend among presumed HIV-negative males aged ≥ 18 years, Washington State, 1996–2007
Figure 3
Gonorrhea rate and incidence trend for HIV-positive males aged ≥ 18 years, Washington State, 1996–2007

DISCUSSION

Significant and persistent inequalities exist in the incidence rate of gonorrhea infections among HIV-positive men compared with their presumed HIV-negative counterparts. Our analyses indicate that, on average, approximately 1.0% of HIV-positive males are being diagnosed with gonorrhea infections each year (range: 0.7% to 2.8%) and account for up to 8.0% (12.3% in 2005) of all male gonorrhea cases diagnosed and reported across the study period in Washington State. These findings suggest ongoing sexual risk-taking among HIV-positive men, which may have potentially serious implications for ongoing HIV transmission. Additionally, differences in the demographic profile of HIV-positive men with gonorrhea compared with presumed HIV-negative men with gonorrhea suggest that older, white MSM are at considerably higher risk of HIV infection, and that when men in this category are being diagnosed and treated for gonorrhea, public health staff may have an important prevention counseling opportunity.

Recent decreases in gonorrhea incidence among both HIV-positive and presumed HIV-negative men are an optimistic development. Yet, the gonorrhea incidence rate among HIV-positive men continues to be almost 20-fold higher than rates observed among HIV-negative men, an inequality persisting across the entire study period. Examination of the trends in annual percent change in rates for HIV-positive and presumed HIV-negative males across the study period (data not shown) indicates that rates changed at roughly the same pace for both groups through 2005. Since then, rates of gonorrhea infection among HIV-positive men have been decreasing more rapidly than among presumed HIV-negative men.

The contribution of HIV-positive cases to the overall burden of male gonorrhea infection at the population level remains significant and has implications for the efficacy of Washington State's public health response to changing gonorrhea incidence, including partner management services. The demographic and behavioral characteristics of HIV-positive men with gonorrhea indicate that the majority of cases are reported among MSM who are also older than presumed HIV-negative, mostly heterosexual males diagnosed with gonorrhea infection. Of interest, too, are the clinical settings where gonorrhea infections are being diagnosed among HIV-positive individuals, especially in light of a recent programmatic focus on opportunities to -integrate services across disease categories. An increasing proportion of co-infected cases are being diagnosed outside of categorical HIV or STD clinical settings.

Limitations

Deterministic matching of case records between separate disease registries has significant limitations. We have adjusted our matching algorithm such that bias is steered in the direction of underestimating co-infection by excluding a number of potentially true matches between the datasets if records failed to match deterministically with a sufficient score on a combination of the selected criteria. Incomplete reporting of gonorrhea and reporting delay for HIV cases also affected these analyses, especially for more recent time periods. Denominators for calculating incidence of gonorrhea among HIV-positive men were obtained directly from surveillance registries and not adjusted for reporting delay or to estimate the number of HIV-positive men who were unaware of their status, which would tend to lower the observed incidence rates in HIV-positive men.

It is also plausible that men receiving care for HIV infection may be more likely than presumed HIV-negative men to be screened for gonorrhea infections. However, a majority of men diagnosed with gonorrhea are likely to experience symptoms, which may mitigate potential screening bias based on increased access to primary care among HIV-positive men. The high proportion of HIV-positive men in this study population reporting MSM or MSM/IDU risk also suggests that our findings may be in part reflecting epidemic levels of gonorrhea among the broader population of MSM, which is a serious concern in its own right. Reliable population-based estimates of the number of MSM are not available to quantify this effect.

Finally, these analyses are based on cases of gonorrhea diagnosed and reported during the study period and not on unique individuals. Re-infection rates during the study period were not assessed; differences in the likelihood of re-infection between HIV-positive and presumed HIV-negative men may affect the observed rate inequality.

CONCLUSIONS

Notwithstanding these limitations, our findings demonstrate clear inequalities in gonorrhea incidence by HIV status that merit continuing public health attention. The potential of co-infection with gonorrhea to synergistically promote HIV transmission is a relevant concern given our finding of disproportionately high incidence rates of gonorrhea among HIV-positive men. Our ongoing ability to monitor population-level trends in co-infection provides public health planners in Washington State with an important tool for more appropriately focusing HIV prevention resources, including prioritizing HIV-positive men diagnosed with STDs for enhanced prevention services and recommending routine HIV counseling and testing for MSM diagnosed with gonorrhea as a standard of care across the full continuum of clinical settings.

Our findings also demonstrate that registry matches can reveal interesting and significant secular trends in co-infection and help to provide an evidence base to direct disease prevention efforts in a timely manner. Methods described in this article may be useful in other jurisdictions with similar separate infrastructures for HIV and STD surveillance. Routine registry matching may also be an important interim step demonstrating the utility of more fully integrating categorical surveillance systems across a broad spectrum of diseases sharing significant comorbidity, exposure risk, or population overlaps. Empirical evidence for sustained rates of HIV and gonorrhea co-morbidity further argues for real-time integration of surveillance systems to help indentify significant increases or aberrations in comorbidity, such as the spike in co-infection observed in 2005 in our data. Timely identification of increases in gonorrhea or other bacterial STDs among HIV-positive individuals may provide a critical window of opportunity to minimize ongoing HIV transmission through intensified partner services.

REFERENCES

1. Fleming DT, Wasserheit JN. From epidemiological 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. [PMC free article] [PubMed]
2. Cohen MS, Hoffman IF, Royce RA, Kazembe P, Dyer JR, Daly CC, et al. Reduction of concentration of HIV-1 in semen after treatment of urethritis: implications for prevention of sexual transmission of HIV-1. AIDSCAP Malawi Research Group. Lancet. 1997;349:1868–73. [PubMed]
3. Zhang J, Li G, Bafica A, Pantelic M, Zhang P, Broxmeyer H, et al. Neisseria gonorrhoeae enhances infection of dendritic cells by HIV type 1. J Immunol. 2005;174:7995–8002. [PubMed]
4. Engering A, Van Vliet SJ, Geijtenbeek TB, Van Kooyk Y. Subset of DC-SIGN(+) dendritic cells in human blood transmits HIV-1 to T lymphocytes. Blood. 2002;100:1780–6. [PubMed]
5. Centers for Disease Control and Prevention (US) Comprehensive HIV prevention: essential components of a comprehensive strategy to prevent domestic HIV, 2006. Atlanta: CDC; 2006. Jan, [cited 2008 Jul 10]. Also available from: URL: http://www.cdc.gov/hiv/resources/reports/comp_hiv_prev/pdf/comp_hiv_prev.pdf.
6. CDC (US) HIV/AIDS statistics and surveillance. [cited 2008 Jul 10]. Available from: URL: http://www.cdc.gov/hiv/topics/surveillance.
7. CDC (US), Division of STD Prevention. Sexually transmitted diseases: STD*MIS. [cited 2008 Jul 10]. Available from: URL: http://www.cdc.gov/std/std-mis.
8. SAS Institute, Inc. SAS®: Version 9.1.2. Cary (NC): SAS Institute, Inc; 2004.

Articles from Public Health Reports are provided here courtesy of SAGE Publications