Sexually transmitted infections (STIs) are a significant problem among poor blacks.1,2 Centers for Disease Control and Prevention surveillance data shows that disparities exist between STI rates in blacks and whites. For example, in 2007, 48% of incident cases of Chlamydia trachomatis infection were reported in blacks, and the rate was 8 times greater than the rate in whites. Seventy percent of the cases of Neisseria gonorrhoeae were among blacks that year, and the rate was 19 times greater than the rate in whites.1 Trichomonas vaginalis prevalence ranges from 1.5- to 10-fold greater in black women than women of other racial/ethnic groups.3 Multiple factors have been implicated as contributors to this phenomenon.
Factors associated with these disparities include both proximate and distal social determinants, which are directly and indirectly associated.4 Proximate determinants are well-studied, which include individual sexual behaviors such as unprotected intercourse and multiple partners.5 More distal factors include sexual networks, marriage rates, incarceration rates,6 – 8 and neighborhood factors, such as neighborhood physical conditions and neighborhood composition.6,9 Blacks have high-risk networks,10 low marriage rates, and high incarceration rates, all of which are associated with increased rates of STIs.6 – 8 They are also more likely than whites to live in poor neighborhoods.
Also implicated in the disparate STI rates experienced by blacks is access to care issues.11,12 Blacks tend to have worse access to care, which is associated with higher rates of several adverse health outcomes. Extended travel time to one's healthcare provider is a component of the concept of geographic healthcare accessibility.13 Limited ability to reach care providers may increase the duration of infection, which increases the time during which they are able to transmit infection to others.
Several articles in the literature conclude that STIs abound in areas where there is limited access to STI treatment, and found that when care is more convenient to access, care seeking delays are shorter.14,15 However, other literature concludes that greater travel time is not linked to care delay or to STI diagnosis.16 Greater travel time has even been linked with decreased STI diagnosis.17 More research is needed to clarify the relationship between travel time and STI infection status in light of these conflicting results.
We hypothesized that length of travel time would be positively related to the likelihood of testing positive for an STI in a sample of black public housing residents scheduled for relocation. Black public housing residents are at high risk for STI18 – 20; however, the role of travel time is not well understood in this population, or in other populations. As housing authorities in cities throughout the United States are participating in the HOPE VI (Housing Opportunities for People Everywhere), an initiative to move residents, the majority of whom are blacks, from areas of concentrated poverty to scattered site housing, now is an opportune time to understand the role of travel time in shaping STI risk. Additional understanding of risk in this group might optimize health outcomes in residents of new environments where travel time to healthcare may or may not change. Ultimately, understanding better if there is a relationship of travel time to STI may help decrease STI health disparities.
MATERIALS AND METHODS
Sample and Recruitment
We recruited a sample of sexually active black adults from November 2008 to June 2009 who were living in 1 of 5 Atlanta Housing Authority (AHA) communities in Atlanta, GA, scheduled for emptying through the HOPE VI initiative. Inclusion criteria were: black, resident of public housing complex scheduled for emptying, >17 years of age, sexually active within the past year, and not living with an individual who was already enrolled in the study. We oversampled alcohol and other drug abusers (AOD abusers). AOD abusers were defined as individuals who had engaged in binge drinking during the reporting period, or used illegal drugs during the reporting period.
Participants were recruited through a variety of methods, including staff outreach at each community; flyers posted inside each community and in local community-based organizations and health clinics; and word of mouth. Collectively, these methods produced a convenience sample of 108 residents of 5 AHA communities. Data were collected in places that permitted privacy; these places included rooms at community organizations, faith-based organizations, and Emory University.
The study was approved by the Emory University institutional review board (Approval number: IRB00008710). Before data collection, participants gave informed consent. Afterward, participants received a $25 cash incentive for their time.
The dependent variable was testing positive for at least 1 of 3 STIs: N. gonorrhoeae, C. trachomatis, or T. vaginalis. Participants were given a urine collection cup and instructed to provide a sample of first stream urine. The presence of N. gonorrhoeae and C. trachomatis in clinical specimens was determined by using the Becton Dickinson Probe Tec ET Amplified DNA Assay (Sparks, MD). The sensitivity/specificity of the CT assay is 92%/96.6%; and the sensitivity/specificity of the GC assay is 95.2%/98.8% (package insert). The presence of T. vaginalis in clinical specimens was determined by using Taq Man PCR. The sensitivity/specificity of the TV assay is 100% and 99.6%.21,22
All other variables were ascertained via survey. Participants completed an interviewer-administered computer-assisted personal interview with audio components. Questions about drug use and sexual activity were queried using audio computer-assisted self-interview to improve the validity of responses.23 The main predictor variable was travel time to the place “where you get most of your care.” We queried several demographic characteristics and sexual behaviors including gender, income, employment, education, condom use, partner concurrency, and sex work. Questions, unless otherwise stated, pertained to the 6-month period before relocation of the first 50% of housing community residents because communities were changing rapidly as more and more residents moved away, and researchers wanted to capture data to reflect a time when the communities were mostly intact.
Descriptive statistics were performed to capture the distribution of each variable. Analyses were run using generalized estimating equations that address clustering within public housing communities. We used the Box-Tidwell method to learn whether a nonlinear relationship between travel time and STI status existed.24 Based on our findings, we squared the travel time variable to linearize its relationship to the outcome. To ensure the relationship between travel time and STI status was robust, we included many covariates in the final model that are known to predict STI status. The final multivariate model controlled for participant sociodemographic characteristics, risky sexual behaviors, prior STI history, condom use, and partnership characteristics. We could not include both household income and educational attainment in the model because of the association between these 2 variables, and so included only educational attainment. Analyses were conducted in SAS version 9.2.
Thirty-four percent of the sample of 108 participants was men (Table 1). Ages ranged from 18 to 59 years and the median household income was $7500. About one half of the sample completed high school or earned a General Equivalency diploma, and <1 in 5 was married.
Over 37% of the sample was found to be positive for at least 1 of 3 STIs. Percent positive for at least one STI ranged from 26% to 53% across public housing communities, and there was variation in travel time across and within communities (Table 2). Thirty-six residents tested positive for 1 or more STI. In the bivariate analysis, gender (being a woman) and age (being younger) were associated with testing positive for at least 1 STI (Table 3). In this sample of deeply impoverished black women and men, there was no relationship between alcohol or other drug dependence and STI status.
Median travel time to the place where participants get most of their care was 30 minutes, and there was variation in travel time across and within communities (Table 2). Women with infection had a pattern of longer travel times compared to women without infection (Fig. 1A). A similar pattern was suggested for men, but was not definitive because of the small number of positive men (Fig. 1B).
The multivariate analysis indicated that there was a curvilinear relationship between travel time and STI status (Table 3, Fig. 2). Specifically, when travel times was <48 minutes, higher travel times were associated with a greater likelihood of testing positive for one or more STIs; when participants reported travel times ≥48 minutes, higher travel times were associated with a lower likelihood of testing positive for 1 or more STIs. Statistically significant relationships between travel time to health care provider and STI in a sample of public housing residents remained when we controlled for covariates including sociodemographic characteristics, risky sexual behaviors and partnership characteristics, and a prior history of an STI. Age, thinking partner has multiple partners, believing partner has an STI, and always using condoms were found to have statistically significant associations with STI when we controlled for other variables.
In this sample of poor black public housing residents with a high prevalence of STIs, we found a curvilinear relationship between travel time and likelihood of testing positive for an STI. We hypothesized that there would be a positive relationship on the basis that STI prevalence would be higher in a sample with poorer access to treatment, and that people with longer travel times have poorer access to care than people with shorter travel times. The data supported our hypothesis in the subset of the sample with travel times <48 minutes. However, when travel times were ≥48 minutes the relationship was inverted.
Findings in previous studies are conflicting; however, they do not show the curvilinear relationship found here. In a qualitative study, Goldenberg found that participants reported that geographic isolation and rigorous work schedules limited access to testing and treatment.15 In contrast, a study of STI clinics across England found that longer travel times were not associated with delays in seeking care for STI, and travel time was not associated with an STI diagnosis.16 In addition, a medical record review of hepatitis C status in France showed that greater travel time or distance to provider was associated with decreased detection of hepatitis C.17 However, the latter finding may be a product of decreased access to screening and not necessarily decreased prevalence. The current study better ascertained STI status because it tested subjects instead of reviewing medical records. In the hepatitis C study, a negative result may signify that people were not tested; however, in the current study, a negative result is more precise in stating that participants lacked infection.
Regarding the proposed mechanism of the longer travel time association with more STI, the first half of the curve (Fig. 1). Mercer et al in a cross-sectional survey of over 3000 patients at 4 STI clinics found there was decreased delay in seeking care when patients were able to have walk-in appointments, suggesting that a form of convenience lends itself to more timely utilization.14 Their study did not look at travel time specifically. The current study is limited in that it does not ask whether travel time contributes to a delay in seeking care, or whether the STI-positive participants were more likely to delay care. We conducted post hoc analyses to understand whether poor access to transportation could help explain the relationship between travel time and STI positivity, and found that while travel time was associated with using public transportation to reach the usual source of care, public transportation use was not associated with STI status. Additional research is needed to clarify the pathway of association.
Given the documented associations between usual source of care and health insurance status to health and healthcare use,25 – 27 we included these 2 variables in the model. Including these 2 variables in the model did not alter the magnitude of the relationship between travel time and STI.
Another possibility is that the negative slope portion of the curvilinear relationship between travel time and STI status was produced by voluntary social isolation. Voluntary social isolation may be adopted by some residents of impoverished high-crime communities as a way of coping with the threat of violence.28 – 30 A subset of the current study sample was interviewed qualitatively and revealed that voluntary social withdrawal is practiced by members of the cohort to minimize exposure to local violence (Wolfgang, 2010). Participants who use this strategy may have chosen to get care farther from home as a part of their social isolation. Voluntary social withdrawal may also translate into less sexual activity—a behavior that would protect against STIs. The current analysis cannot explore this possibility because we it did not query voluntary social isolation on the quantitative survey.
Our study has several limitations. Unlike STI status, the travel time variable was gathered by self-report, which is subject to recall bias. Participants may over- or underestimate the amount of time it takes to reach their source of health care. Another limitation is that we examine prevalence of STI at one specific point in time. A more robust result would have come if the study were longitudinal and had tested for STIs at several time points to create a composite score. Furthermore, although STI test results reflect the time of the interview, travel time answers to questions referred to an earlier time. It is not expected that participants would have changed their location of health care before changing residences, but this is a possibility, and if they had, then travel times at the time of interview may have been different. We do not know if the usual source of care reported by our participants is their source of STI-related care or reproductive health care nor do we know the number of healthcare visits per year. There was no relationship between travel time and having received primary care in the past 6 or 12 months. Our sample is a convenience sample and may not be representative of housing authority populations in Atlanta, GA. However, a comparison of our sample's sociodemographic characteristics with those of the underlying population of residents of the 5 AHA communities indicates that the sample is similar to the underlying resident population on all dimensions for which data are available (i.e., age, gender, and income) from the US Department of Housing and Urban Development.
Results of this analysis raise the question of whether people should be encouraged to get care from providers who require the least travel time. Olonilua et al found that only 7.5% of new GUM clinic attendees attended their nearest clinic.16 People chose to travel further to get care. On the other hand, Andersen et al found, in a sample of low income persons, that those living in areas where there were more federally funded health centers were more likely to get medical care than those residents of areas with fewer federally funded health centers.31 It will be important to do qualitative research to see if public housing residents who get their care from providers who are farther away differ systematically from those who get their care from nearby clinics. For example, are the longer travel time participants less socially connected to their neighbors? Do they separate themselves from local sexual networks that may have high background STI prevalence and form partnerships with individuals living in neighborhoods where STIs are not so highly prevalent?
STIs are health problems plaguing poor black disparately. Residents of impoverished communities experience improved sexual health if they access services that require less travel time to reach. However at longer travel times, the relationship is inverted and this overall curvilinear relationship has not been reported in previous literature. Qualitative studies are important to help reveal the possible causes of this phenomenon. In the meantime, STI testing and treatment facilities ought to be accessible especially to vulnerable groups in an effort to decrease health disparities.
1. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance. 2008. U.S. Department of Health and Human Services, Atlanta, Georgia; 2009.
2. Krueger LE, Wood RW, Diehr PH, et al.. Poverty and HIV seropositivity: The poor are more likely to be infected. AIDS 1990; 4:811–814.
3. Shafir SC, Sorvillo FJ, Smith L. Current issues and considerations regarding trichomoniasis and human immunodeficiency virus in African-Americans. Clin Microbiol Rev 2009; 22:37–45.
4. Link BG, Phelan J. Social conditions as fundamental causes of disease. J Health Soc Behav 1995; Spec No:80–94.
5. Holmes KK, Sparling PF, Stamm WE, et al., eds. Sexually Transmitted Diseases. 4th ed. New York, NY: McGraw-Hill, 2008.
6. Cohen DA, Mason K, Bedimo A, et al.. Neighborhood physical conditions and health. Am J Public Health 2003; 93:467–471.
7. Thomas JC, Sampson LA. High rates of incarceration as a social force associated with community rates of sexually transmitted infection. J Infect Dis 2005; 191(suppl 1):S55–S60.
8. 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.
9. Latkin CA, Curry AD, Hua W, et al.. Direct and indirect associations of neighborhoo disorder with drug use and high-risk sexual partners. Am J Prev Med 2007; 32(suppl):S234–S241.
10. Liljeros F, Edling CR, Amaral LA. Sexual networks: implications for the transmission of sexually transmitted infections. Microbes Infect 2003; 5:189–196.
11. Parrish DD, Kent CK. Access to care issues for African American communities: Implications for STD disparities. Sex Transm Dis 2008; 35(suppl):S19–22.
12. Newman LM, Berman SM. Epidemiology of STD disparities in African American communities. Sex Transm Dis 2008; 35(suppl):S4–S12.
13. Aday LA, Andersen R. A framework for the study of access to medical care. Health Serv Res 1974; 9:208–220.
14. Mercer CH, Sutcliffe L, Johnson AM, et al.. How much do delayed healthcare seeking, delayed care provision, and diversion from primary care contribute to the transmission of STIs? Sex Transm Infect 2007; 83:400–405.
15. Goldenberg SM, Shoveller JA, Ostry AC, et al.. Sexually transmitted infection (STI) testing among young oil and gas workers: The need for innovative, place-based approaches to STI control. Can J Public Health 2008; 99:350–354.
16. Olonilua O, Ross JDC, Mercer C, et al.. The limits of health-care seeking behaviour: how long will patients travel for STI care? Evidence from England's Patient Access and the Transmission of Sexually Transmitted Infections (PATSI) study. Int J STD & AIDS 2008; 19:814–816.
17. Monnet E, Ramee C, Minello A, et al.. Socioeconomic context, distance to primary care and detection of hepatitis C: A French population-based study. Soc Sci Med 2008; 66:1046–1056.
18. Sikkema KJ, Koob JJ, Cargill VC, et al.. Levels and predictors of HIV risk behavior among women in low-income public housing developments. Public Health Reports (Washington, DC, 1974). 1995; 110:707–713.
19. Sikkema KJ, Heckman TG, Kelly JA, et al.. HIV risk behaviors among women living in low-income, inner-city housing developments. Am J Public Health 1996; 86:1123–1128.
20. Sikkema KJ, Heckman TG, Kelly JA. HIV risk behaviors among inner-city African American women. The Community Housing AIDS Prevention Study Group. Women's Health (Hillsdale, NJ) 1997; 3:349–366.
21. Ingersoll J, Bythwood T, Abdul-Ali D, et al.. Stability of Trichomonas vaginalis
DNA in urine specimens. J Clin Microbiol 2008; 46:1628–1630.
22. Caliendo AM, Jordan JA, Green AM, et al.. PCR improves detection of Trichomonas vaginalis
infection compared with culture using self-collected vaginal swabs. Infect Dis Obstet Gynecol 2005; 13:145–150.
23. Rogers SM, Willis G, Al-Tayyib A, et al.. Audio computer assisted interviewing to measure HIV risk behaviours in a clinic population. Sex Transm Infect 2005; 81:501–507.
24. Hosmer DH, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley and Sons, 1989.
25. Ayanian JZ, Weissman JS, Schneider EC, et al.. Unmet health needs of uninsured adults in the United States. JAMA 2000; 284:2061–2069.
26. Corbie-Smith G, Flagg EW, Doyle JP, et al.. Influence of usual source of care on differences by race/ethnicity in receipt of preventive services. J Gen Intern Med 2002; 17:458–464.
27. Weissman JS, Epstein AM. Falling through the safety net: Insurance status and access to health care. Baltimore, MD: Johns Hopkins University Press, 1994.
28. Skogan W. Fear of crime and neighborhood change. Crime Justice 1986; 8:208–229.
29. Furstenberg FF. How Families Manage Risk and Opportunity in Dangerous Neighborhoods. Newbury Park, CA: Sage Publications, 1993.
30. Riger S, Gordon MT, LeBailly RK. Coping with urban crime: women's use of precautionary behaviors. Am J Community Psychol 1982; 10:369–386.
© Copyright 2012 American Sexually Transmitted Diseases Association
31. Andersen RM, Yu H, Wyn R, et al.. Access to medical care for low-income persons: how do communities make a difference? Med Care Res Rev 2002; 59:384–411.