Individual and population-level factors associated with Chlamydia trachomatis infection among men who have sex with men screened in Washington State Infertility Prevention Project clinics, 2003-2011

Charles Jack Shumate, The University of Texas School of Public Health


Background: Chlamydia trachomatis (CT) is an extremely common, but underestimated STD since the majority of infections are asymptomatic and screening protocols rely on client disclosure of risk factors in the absence of symptoms. Moreover, CT disproportionately affects racial minority sub-group, and a majority of these infections occur among men who have sex with men (MSM). Current CT risk-based screening protocols have failed to address the epidemic among MSM, and there are gaps in the literature preventing the estimation of pertinent risk factors associated with CT infection among MSM e.g. studies of usually symptomatic males, the lack of sexual orientation or partner preference measure, and the absence of population-level factors. This dissertation examined the relationship between individual and population-level factors associated with CT infection among an asymptomatic, screened population of MSM in Washington State participating Infertility Prevention Project (IPP) Family Planning/Reproductive Health (FP/RH) and STD clinics, 2003-2011.^ Methods: IPP clinic screening level data of records where ‘male’ sex and client self-report of ‘only sex with other men’ were assembled into a dataset (n=284, 841, records). Diagnostic tests, i.e. MSM with signs, symptoms indicating a current STD infection or report of exposure to an STD from a sex partner were excluded. A total of 9,645 records met the inclusionary criteria of an asymptomatic MSM record. Chapter 2 (paper 1) presents the results of a cross-sectional descriptive epidemiologic study of CT positivity of asymptomatic males (n=9,645, records) screened in WA State FP/RH and STD clinics, 2003-2011. Frequency distributions and cross-tabs were computed for demographics, behavioral risk factors, specimen site and CT test result. With CT positivity as the dependent variable, unadjusted incidence rate ratios [IRR] IRR ratios for each independent variable were calculated from a univariate regression model. All variables were included in a simultaneous modified Poisson regression model with robust variance estimation to obtain adjusted IRRs. The IRRs from the modified Poisson regression were evaluated along with each variables 95% CI and the robust standard errors. Interaction between race and behavioral risk factors was assessed. Chapter 3 (paper 2) presents the results of a retrospective (n=823, unique clients) time-to-event study of incidence of CT infections and repeat infections, and the associations CT infections and demographic, behavioral and population-level risk factors among a sample of MSM screened in one WA State STD IPP clinic. The demographic and individual risk factors with the addition of population-level factors (Median household Income, the percentage of the population 100% below the Federal Poverty Level [FPL] and the percentage of the population comprising racial minority groups by client residential ZIP code) were included in the study to assess their role in increasing risk for infections in racial sub-groups of MSM. Cox regression analysis was used to estimate the adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for repeat CT infection, according to the different risk factors. Chapter 4 (paper 3) presents the results of a retrospective spatiotemporal exploratory study which assessed ZIP code geographic clustering of CT infections (n=798, positive CT cases) among clients screened in IPP FP/RH and STD clinic, 2003-2011. For this study, I described the geographic characteristics of clustering for CT cases by quintiles of median household income and percentage of population comprising racial minority groups, and carried out a cluster analysis using the space-time permutation scan statistics (STPSS), implemented in the SaTScan software program. The STPSS was used to identify when and where CT infections occurred at higher levels than expected by chance.^ Results: Paper 1 found CT positivity was high among MSM screened in Washington State (8.96%), and positivity varied by venue of testing 9.6% and 8.7% in FP/RH and STD sites, respectively. The factors of younger age 15-17 overall and younger age 15-17 in FP settings maintained statistical significance for reduced risk for CT infection in univariable and multivariable models. By race/ethnicity, only Blacks and Hispanics maintained statistical significance for increased risk overall for combined non-FP and FP settings; by setting, non-Hispanic Blacks in FP had increased risk for CT infection, while CT positivity among clients attending STD clinic were Asian/Pacific Islander and Hispanic Status. Paper 2 demonstrated that while overall CT prevalence was high (∼8%), repeat CT infections were low (∼3%) in the sample of MSM limiting the ability to measure the associations between risk factors and repeat infections. The associations between incident infections and risk factors were measured Non-whites had worse CT-specific survival than whites. The risk for CT infection among non-whites increased after adjusting for covariates and maintained statistical significance. CT prevalence decreased as area SES improved, with the highest positivity found in ZIP codes with the highest median household income, but reverse for positivity and lowest percentage of the population living in poverty; however, the population-level factors were not statistically significant in the overall model. Paper 3 revealed that there were 798 CT infections among MSM during the 1/1/2003- 12/31/2011 study period across 170 ZIP codes and 54 participating IPP STD and FR/RH clinics. The most likely cluster for high CT was found in 8 client ZIP codes for the period 1/7/2004 to 12/31/2005 using a 9 month time window and again in the 12 month time window.^ Discussion: This study adds critically needed information to the existing knowledge base that will inform future research on CT control among MSM. The first journal article fills a gap in the scientific literature regarding whether individual CT risk factors vary by racial sub-group of MSM asymptomatically screened. Current guidelines for CT screening focus primarily on “high risk” MSM. This strategy may require reconsideration, given the increased risk of infection among MSMs overall and by racial sub-group after controlling for individual risk. The second journal article fills a gap in the literature regarding the incidence of repeat CT infections among MSM and the role of population-level factors, such as residential racial segregation, have on increase CT risk among MSM. Universally screening MSM based on select criteria may be of public health benefit, especially if the appropriate groups are targeted such as MSM with repeat infections or by ZIP codes with population-level attributes associated with increased infection risk. The third journal article fills a gap in the literature regarding the geographic clustering of CT among MSM. Preliminary research suggested the spatial clustering of CT among males in general, and specifically among MSM, was weak; however, CT clustering has not been investigated sufficiently among MSM. This study demonstrated that risks for CT are associated with geographic spaces (ZIP codes) and provides guidance for targeted interventions to supplement existing CT control measures aimed at the general population. Locating geographic areas of increased risk and targeting them for increased intervention is the primary method to decrease STDs within the population.^

Subject Area

LGBTQ studies|Epidemiology

Recommended Citation

Shumate, Charles Jack, "Individual and population-level factors associated with Chlamydia trachomatis infection among men who have sex with men screened in Washington State Infertility Prevention Project clinics, 2003-2011" (2016). Texas Medical Center Dissertations (via ProQuest). AAI10183277.