Levels of chlamydia among 15–24-year-old women screened through a federal program in the Northwest in 1997–2006 were linked more strongly to characteristics of the women themselves than to characteristics of the areas in which they lived.1 Throughout the period, tests among white women were less likely than those among other racial or ethnic groups to detect infection. Within each racial and ethnic group, the likelihood of infection was associated with women’s age, certain sexual risk behaviors and STD history; associations with other individual characteristics were significant in some, but not all, groups. By contrast, socioeconomic characteristics of women’s neighborhoods (broadly defined by residential zip code areas) showed relatively few associations with the risk of infection, and findings varied across groups.
The U.S. Public Health Service Region X Infertility Prevention Project provides chlamydia screening and treatment for women attending family planning clinics in Alaska, Idaho, Oregon and Washington; it recommends routine testing for all female clients younger than 25. Between 1997 and 2006, roughly 700,000 tests were performed. To assess characteristics that are associated with chlamydia infection, researchers examined individual-level data from clinic records and interviews women completed at the time of a clinic visit, as well as aggregate-level information (e.g., the proportion of a neighborhood’s population that had a given characteristic) obtained from 2000 census data. They calculated chlamydia positivity and used univariate and multivariate logistic analyses to identify indicators of risk. Aggregate-level measures were coded as ordered categories, and the lowest category of each was used as the reference group in the analyses.
Five percent of tests over the study period revealed chlamydia infection. Tests for white women were significantly less likely than those for other racial and ethnic groups to have positive results (5% vs. 6–10%), a finding confirmed in analyses controlling for women’s characteristics (odds ratios, 1.3–1.9). Data for individual years show a similar pattern throughout the period. However, because the level of infection increased steadily over time among whites, but not among other groups, the gaps narrowed.
At the univariate level, age, urban-rural residence and several measures of sexual behavior and STD history also were related to the likelihood of infection. Notably, the risk was considerably elevated among women who visited the clinic because they had a sex partner who had chlamydia, women who reported having had a symptomatic partner in the last 60 days, women whose clinical exams revealed findings consistent with an STD and women who had tested positive for chlamydia in the past year (odds ratios, 2.4–6.5). The odds of infection were inversely related to the median household income of the areas in which women lived (0.9 for those in the highest income quintile) and were positively related to the proportions of the population who lived in poverty, were adults aged 25 or older who had not graduated from high school, who belonged to racial minority groups and who were Hispanic (1.1–1.8 for the highest categories of these measures).
Results of multivariate analysis confirmed that within every racial or ethnic group, the risk of chlamydia infection was elevated among teenagers (odds ratios, 1.2–1.7), women who visited the clinic because their partner had chlamydia (1.9–4.2), those who had recently had a symptomatic partner (1.9–3.1) and those whose clinical exams detected possible signs of an STD (2.6–3.5). Findings for other individual characteristics were less consistent. For example, all women except Hispanics had an elevated risk if they had had a new partner in the past 60 days (1.3–1.5), and all except American Indians and Alaska Natives were at increased risk if they had had a positive chlamydia test in the past year (1.2–1.8). For white women and Asians and Pacific Islanders, risk was negatively associated with condom use at last sex (0.9 for both).
Household income of women’s area of residence was not associated with infection risk, but for Hispanics, risk increased with the proportion of the population who were poor (odds ratio, 1.1). Among whites, risk was positively associated with the proportion of adults who lacked a high school diploma (1.1), but among Hispanics, the opposite was the case (0.9). For black women, risk increased with the proportion of the population who belonged to racial minority groups, and for Hispanics, it rose with the proportion who were Hispanic (1.1 for each).
While acknowledging that the project (like similar ones in other regions of the country) "captures a limited array of client risk indicators," the researchers comment that their results "point toward further work needed … on client sexual partner and network characteristics, as well as other social determinants [of chlamydia infection]." Furthermore, because the types of aggregate-level measures they used are available in most regions, the researchers suggest that such measures could play a role in program planning.—D. Hollander