Plitt, Sabrina S. PhD*; Garfein, Richard S. PhD†; Gaydos, Charlotte A. DrPH‡; Strathdee, Steffanie A. PhD§; Sherman, Susan G. PhD*; Taha, Taha E. PhD*
STUDIES CONSISTENTLY DEMONSTRATE THAT injection drug users (IDUs) practice sex behaviors that place them at high risk for sexually transmitted infections (STIs).1,2 However, information on nonbloodborne STIs among drug users is sparse, and most studies are based on a small number of IDUs captured in samples from clinics or drug treatment programs, which typically represent older, longer-term injectors. Recent studies of younger IDUs involving community-based samples show a high prevalence of sexual practices that could expose them to STIs.2,3 One study found that even among young IDUs who had a high prevalence of unprotected sex, the prevalence of gonorrhea and chlamydia were relatively low2; however, little is known about the prevalence of other STIs in this population.
The public health consequences of STIs are heightened among IDU populations as a result of the potentiation of HIV transmission and acquisition by both ulcerative and nonulcerative STIs.4–8 These consequences of STI–HIV coinfections are important among populations such as IDUs in whom the prevalence of HIV infection is greater than the general population and the probability of having HIV-positive sex partners is increased. In this study, we attempted to confirm a low prevalence of chlamydia and gonorrhea, estimate the prevalence of trichomoniasis and bacterial vaginosis (BV), and identify correlates of these infections among a cohort of young IDUs in Baltimore, Maryland.
Materials and Methods
Young adult IDUs were recruited for participation in this study from the Drug Users Intervention Trial (DUIT), a multicenter, randomized, controlled trial evaluating an intervention of cognitive–behavioral skills building to reduce sexual and injection risks associated with HIV and hepatitis C virus (HCV) infection. DUIT recruitment was conducted using community-based outreach, targeted sampling, and peer-driven sampling.9 The baseline assessment visit was designed to provide cross-sectional data on prevalence of multiple infections among young IDUs; therefore, recruitment and screening activities were conducted without regard for prevalent infection status. Participants were deemed eligible for DUIT if they were 15 to 30 years old, were not enrolled in other studies at the same time, planned to be in the area for the next 12 months, and had injected at least once in the 6 months before enrollment. All eligible participants were asked to complete a baseline survey and underwent HIV and HCV antibody screening with appropriate pretest counseling according to Centers for Disease Control and Prevention (CDC) guidelines.10
At the Baltimore site, between September 2002 until February 2003, all eligible participants for DUIT aged 18 to 30 years were subsequently asked if they wanted to join the STI substudy. Participants who agreed proceeded with informed consent and baseline assessment. At the end of the visit, participants were reimbursed for their time and travel costs, $20 for DUIT and $5 for the STI substudy. This research was approved by the Committee on Human Research at the Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, and the Institutional Review Board at the CDC, Atlanta, Georgia.
All participants were asked to complete the DUIT baseline questionnaire and an additional STI study questionnaire. The DUIT questionnaire was administered by audio computer-assisted self-interview (ACASI), which has been associated with higher reporting levels of sensitive behaviors among IDUs,11,12 as well as greater validity.13 In addition to demographic information, this questionnaire gathered information on sexual and drug use behaviors such as age at first intercourse, number of sex partners, frequency of condom use with steady or casual partners, type of sex (vaginal/oral/anal), type and frequency of drug use, self-reported description of sex partners, homosexual/bisexual behaviors, history of trading sex for money or drugs, and the frequency of “ask[ing] a sex client… to use a condom.” The majority of the sexual and drug use behavior variables referred to the previous 3-month period. A main steady sex partner referred to the sex partner to whom the participant felt was their closest or most important sex partner. Casual partners were defined as “nonsteady” or sex-trading partners.
Participants were also asked to complete a 1-page interviewer-administered questionnaire to collect information that was not captured by the DUIT baseline questionnaire such as lifetime sexual partners and history of STIs. In addition, douching practices of female participants were assessed by questions referring to douching frequency in the previous 6 months, date of last douche, reasons for douching, and type of douching solution used.
Two self-administered vaginal swabs were collected from women and 1 “first-catch” urine specimen was collected from men. Specimens were tested for chlamydia, gonorrhea, trichomoniasis, and, among women only, bacterial vaginosis as described subsequently.
Strand displacement amplification (SDA) was used to test for the presence of Chlamydia trachomatis and Neisseria gonorrhoeae DNA in urine samples and self-administered vaginal swabs. The ProbeTec ET system was used following the instructions provided by the manufacturer (Becton Dickinson, Sparks, MD). The presence of Trichomonas vaginalis DNA was tested by polymerase chain reaction (PCR) in urine and in self-administered vaginal swabs. This assay used 2 β-tubulin gene primers, BTUB-2 and BTUB-9.14 Previous studies have shown both the BD ProbeTec ET and the T. vaginalis PCR assay to have high sensitivity and specificity.14–16
Female participants were also tested for BV. Immediately after specimen collection, study staff rolled 1 of the vaginal swabs onto a microscope slide, fixed the smear, and allowed it to air dry. Once transported to the lab, these slides were Gram-stained, examined microscopically by trained laboratory staff, and given a score of zero to 10 based on the relative proportion of BV-associated organisms (i.e., large Gram-positive [Lactobacillus], small Gram-negative or Gram-variable rods [Gardnerella or Bacteroides], and curved Gram-variable rods [Mobiluncus]).17 Slides with a score of 7 or higher were diagnosed as BV, whereas scores of less than 4 were diagnosed as negative for BV. Any slide with a preliminary intermediate reading (i.e., a score of 4–6) was read by a second reader; the final BV score was determined when both readers came to a consensus. Compared with Amsel’s clinical criteria, the sensitivity and specificity of the Nugent criteria is 89% and 83%, respectively.18 In addition, the Nugent criteria have been shown to have high intra- and interobserver reliability.19
Experience shows that it can be difficult for IDUs to return for laboratory test results. To minimize participant burden and maximize the number of participants who would be given their test results, participants were asked to return 2 weeks after baseline to receive their results and posttest counseling for all tests in 1 visit. If positive for any STI or BV, participants were referred for medical evaluation and treatment, if required. A combination of reminder letters, telephone calls, and home visits were used to ensure that participants returned to receive their test results. Participants received monetary compensation for attending this results visit. As required by Maryland State Law, all cases of gonorrhea and chlamydia were reported by name, age, sex, and address to the Baltimore City Health Department, with the participant’s prior consent.
Prevalence was defined as the proportion of subjects with positive tests at their baseline visit. A 95% binomial confidence interval was calculated for each prevalence estimate. Given the observed differences in prevalence by gender, it was decided to stratify all further analyses by gender.
As a result of the small number of infections, data were collapsed for chlamydia, gonorrhea, or trichomoniasis and correlates were not identified for individual STIs. A separate analysis was carried out to assess correlates for BV. Descriptive statistics (e.g., chi-square, Fisher exact tests, Mann-Whitney tests) were used to analyze the distribution of categorical and continuous variables (e.g., demographics, sexual behaviors, drug use behaviors) between those infected and those not infected.
Multivariate logistic regression analyses were performed separately for men and women to identify significant independent effects of demographic, sexual behavior, and drug use variables on being positive for at least 1 of the STIs (i.e., chlamydia, gonorrhea, or trichomoniasis). Any variable associated with the outcome at P <0.15 in univariate analyses was entered into a logistic regression model to obtain adjusted estimates of the odds ratios (AOR) and 95% confidence intervals (CI). Race and age were assessed as possible confounders. If no meaningful changes were noted in the coefficients or the model, these variables were not included. Potential interactions were examined by adding product terms to the logistic regression model. Variables with zero cells could not be included in the logistic regression model. Statistical significance was determined using a significance of 0.05. All analyses were performed using Stata Version 7.0 (Stata Corp., College Station, TX).
Demographics, Drug Use, and Sexual Behaviors
Of 358 eligible individuals invited to join the STI substudy, 340 (95%) agreed to participate. Of the participants who consented, urine or vaginal swabs were received and tested for 326 (96%). Of the remaining 4%, specimens were either not collected from participants during their visit or were not properly transported for testing by the laboratory (e.g., leaked during transport).
Of these 326 participants, 115 (35.3%) were women and 211 (64.7%) were men (Table 1). In total, 75.6% of the sample was white, 14.4% were black, 2.2% Hispanic, 2.2% American Indian, 0.3% Asian, and 5.5% mixed races. The median age of the sample was 24 years (interquartile range [IQR] = 22–27). Twenty-five percent reported being homeless at least part of the time in the last 6 months. The seroprevalence of HIV was 1.7% among females and 2.4% among males.
In the previous 3 months, 71.7% of women and 62.7% of men injected daily (P = 0.10). For both sexes, the predominant drug used was heroin, whereas lower proportions reported injection of cocaine alone or with heroin (i.e., speedball). Crack smoking was reported by over half (53.7%); however, daily crack smoking was only reported for 2.2% of the population and did not differ by gender.
The median age of sexual debut of the sample was 14 years. Men had a higher median number of lifetime sexual partners than women (20 vs. 14.5, respectively; P = 0.02). Women were more likely than men to report bisexual/homosexual orientation (36.6% vs. 7.4%, respectively; P <0.001) and same-sex relationships in the past 3 months (27.3% vs. 10.6%, respectively; P <0.001).
Nearly all participants reported some sexual activity during the past 3 months. Women were significantly more likely to report having a steady sex partner in the 3 months before study entry than men (91.1% vs. 81.4%, respectively; P = 0.02). Only 13% of those with a main steady sex partner and 39.3% of those with casual partners reported consistent condom use for vaginal sex; these proportions did not differ significantly by gender. Women were more likely than men to report having a main steady sex partner who had ever injected drugs (74.0% vs. 42.7%, respectively; P <0.001). In the 3 months before study entry, 35.4% of women and 11.9% of men (P <0.001) had received money or drugs for sex, whereas 6.3% of women and 9.1% of men (P = 0.52) had paid someone to have sex. Women were more likely than men to consistently ask their sex trade partners to use condoms (70.7% vs. 36.7%; respectively; P = 0.004).
Sexually Transmitted Infection Prevalence
In total, 16.4% of women tested positive for chlamydia, gonorrhea, or trichomoniasis. In contrast, 5.2% of men tested positive for any of the 3 infections. In women, the prevalence of chlamydia, gonorrhea, and trichomoniasis was 5.3% (95% CI, 2.0–11.1%), 3.5% (95% CI, 0.95–8.6%), and 8.6% (95% CI, 4.2–15.3%), respectively (Table 2). Only 1 woman was coinfected with chlamydia and gonorrhea. The prevalence of chlamydia and trichomoniasis among men was 3.3% (95% CI, 2.3–6.6%) and 1.9% (95% CI, 0.5–4.7%). Among male participants, there were no cases of gonorrhea. The proportions of participants infected with gonorrhea and trichomoniasis were statistically higher among women than men. In addition, a higher proportion of women than men self-reported a previous infection with chlamydia (19.4% vs. 4.4%, P <0.001) and trichomonas (15.7% vs. 0.5%, P <0.001), respectively (data not shown).
All gonorrhea cases were among white participants, but there was no statistical difference in chlamydia prevalence by race. Among trichomoniasis cases, a significantly higher proportion of cases occurred in black than white participants (12.8% vs. 3.3%, P = 0.02). There were no significant differences in age between cases and noncases for any of the individual infections.
Sexually Transmitted Infection Correlates
Statistically significant correlates for being infected with any STI (chlamydia, gonorrhea, or trichomoniasis) were identified by univariate analyses for men and women separately. The median age was higher among women with at least 1 STI (25 years, IQR = 22–28 years) than among women with no STIs (23 years, IQR = 20–27 years), but this difference was not statistically significant (P = 0.08). The median age for men was 25 years in both groups (P = 0.23). The median age of first sexual intercourse was 14 years overall and did not differ by gender or STI infection status. There was also no difference in the median number of lifetime sexual partners between those with and without STIs among women (13.5, IQR = 5–30 vs. 14.5, IQR = 9–29; P = 0.38) or men (15, IQR = 11–20 vs. 20, IQR = 10–40; P = 0.42).
Women participants were less likely to be infected with at least 1 STI if they reported being in a same-sex relationship (0% vs. 32.3%, P = 0.006) and having a main steady sex partner who injected drugs (35.3% vs. 72.3%; P = 0.005) (Table 3). A greater proportion of women with STIs had douched in the past 6 months than women without an STI (81.3% vs. 56.5%; P = 0.09), although this difference was not statistically significant. This difference was greatest among those who had douched at least 6 times in the past 6 months. No other variables investigated for women were statistically significant. Among men, those with at least 1 STI were more likely to report having had anal sex with a female sex partner in the past 3 months than those with no STIs (72.7% vs. 31.0%, P = 0.007). No other variables investigated for men were statistically significant.
Multivariate analyses of women identified that having a steady sex partner who injected drugs (AOR, 0.3; 95% CI, 0.08–0.8) was an independent correlate for not being infected with at least 1 of the STIs (i.e., gonorrhea, chlamydia, or trichomoniasis), whereas having douched in the previous 6 months (AOR, 4.9; 95% CI, 1.1–23.6) was positively associated with infection. Among men, anal sex with female partners was positively associated with infection (AOR, 6.3; 95% CI, 1.5–25.8), whereas alcohol use was negatively associated with infection (AOR, 0.3; 95% CI, 0.07–1.0).
Bacterial Vaginosis: Prevalence and Correlates
Of the 112 female participants who had an BV smear, 63 (56.3%) were positive for BV, 6 (5.4%) were intermediate, and 43 (38.4%) were negative. Douching was the only factor found to be more common among those women with positive (or intermediate scores) BV smears than among those with negative smears (Table 4). Compared with women without BV, women with BV were more likely to have douched in the previous 6 months (67.2% vs. 44.7%; OR, 2.5; 95% CI, 1.1–5.7). Among women who douched in the previous 6 months, the median number of douches was higher (6 times [IQR = 6–12] vs. 4 times [IQR = 2–6]; OR, 1.1; 95% CI, 1.0–1.3) and the time since last douche was shorter (22 days [IQR = 6–34] vs. 34 days [IQR = 21–62]; OR, 0.98; 95% CI, 0.96–0.99) among those women with BV than those without BV. There was no difference between these groups in terms of type of douching solution (i.e., store-bought product or vinegar solution). No significant differences were found for douching variables after adjustment by age or race.
In this cohort of young IDUs, we observed high levels of sexual risk, including having multiple sex partners and low condom use, yet the prevalence of chlamydia, gonorrhea, and trichomoniasis were moderate to low. The high prevalence of sexual risk behaviors and BV among female participants is of concern as a result of the increased risk of HIV acquisition among women with BV infection.20
The prevalence of chlamydia and gonorrhea reported in this cohort was similar to those reported from other young drug-using populations.2,21–24 In a study of young IDUs by Latka et al. in the CIDUS-II multicenter study, the prevalence of chlamydia was 5.2% and did not differ by gender. Gonorrhea prevalence was 2% among women and 0.2% among men.2 Poulin et al. examined the prevalence of STIs among 738 drug users attending an STI prevention and needle exchange program in Quebec City, Canada. In that study, the prevalence of chlamydia was 5.3% for women and 3.3% for men; gonorrhea was 3.4% among women and 0% among men.24
In our study, there were no cases of chlamydia or gonorrhea among black women. These results were similar to those of a study conducted among female detainees in Baltimore city prisons in which white participants had a significantly higher proportion of chlamydia and gonorrhea infections than did black inmates.25 The black women in our study were significantly older than the white women, which may have led to our finding different prevalence by race, because both of these infections are associated with younger age.26 The prevalence of trichomoniasis was lower than that reported in other drug-using populations. Bachmann et al.21 reported a prevalence of 22% among women attending a residential drug treatment center in Alabama, whereas Lally et al.23 showed a prevalence of 43% among women at a substance abuse treatment center. We are unaware of any published reports of trichomoniasis prevalence conducted specifically among male drug users; however, research among men in Baltimore city sexually transmitted disease clinics (of which 7% were IDUs) indicated that 13% tested positive by urine PCR and/or culture.27 The lower prevalence of trichomonas infection reported from our sample compared with other studies is surprising considering that the use of PCR should have increased sensitivity to detect infections.14 One possible difference is that in those other studies, the populations were older than our participants; older age has been positively associated with trichomonas infection.27,28
Overall, the prevalence of STIs from this population of young IDUs is comparable to those from the general population. A 2002 Baltimore population-based study of 18 to 35 year olds showed that the prevalence of untreated chlamydia and gonorrhea was 3.0% and 5.3%, respectively, with the prevalence being higher among blacks than whites.29 A recent national survey of young adults reported a prevalence of chlamydia and gonorrhea of 4.7% and 0.4%, respectively, among females, and 3.7% and 0.4%, respectively, among males.30 The prevalence of trichomoniasis among women in the general population has been reported to be 5% to 10%, whereas the prevalence among asymptomatic men usually falls below 5%.31
In our study, the prevalence of chlamydia, gonorrhea, and trichomoniasis were all higher among women than men, a pattern seen in similar studies.22,24 Some of the other STI correlates such as decreased alcohol use and not having a sex partner who has injected drugs are difficult to explain; however, given the small number of infections and the large number of variables examined, some of these associations may have occurred by chance. The small sample size, as well as the combination of all STIs into a single outcome variable, may also explain our inability to identify correlates of STIs usually identified in the literature such as exchanging sex for money or drugs and precluded our ability to identify correlates that were specifically related to 1 of the STIs under study.
This sample had a high BV prevalence that was significantly associated with douching. The only other published report of BV prevalence among drug users indicated a similarly high percentage of 61%.21 The implications of BV are important among a population with high risk of HIV infection such as this one, because BV has been observed to potentiate the transmission of HIV.6–8 In addition, BV has also been associated with pelvic inflammatory disease and ectopic pregnancy.32,33 The association of BV with vaginal douching has been shown in other studies34–36; it is hypothesized that douching may upset the natural balance of genital flora thus leading to BV. One limitation of our cross-sectional design was the inability to establish temporality between douching and BV acquisition. However, when reasons for douching were examined, few women douched to relieve vaginal symptoms and BV was not associated with douching to relieve vaginal symptoms (1.5% vs. 5.3%, P = 0.30). Other correlates of BV described in the literature such as black race34 or use of oral contraceptives35 were not observed here. In our study, there was no statistically significant difference in the proportion of black women having douched in the previous 6 months (13 of 18: 72.2%) as compared with white women (47 of 82: 57.3%; P = 0.30), which could help explain why BV prevalence was not found to be associated with race. BV has also been associated in previous literature with the acquisition of a new sexual partner37; however, this variable could not be assessed in the current study. Further research of genital infections among female drug users should include assessments of BV.
In our study population, 60% of the women had douched over the past 6 months. This percentage is high given that the 1995 National Survey for Family Growth indicated that 27% of U.S. women of reproductive age douched38 and that black women, a group commonly associated with douching, represented only 17.4% of the women in our study population. Although research has not established a causal relationship between BV and douching, the association seen in this research and other literature34–37 suggests that further research should assess the importance and implications of counseling on the reduction of douching for female reproductive health.
This study was limited by low statistical power as a result of the low prevalence of infection, which led to difficulty identifying significant correlates. The baseline questionnaire did not collect information on sexual networks and sexual mixing patterns (e.g., concurrency) that have been shown to be highly correlated with STI risk.39 In addition, the questionnaire did not collect information on existing STI symptoms; however, any participant reporting symptoms during pre- or posttest counseling was referred for medical evaluation. All behavioral questions were limited to the previous 3-month period, which may have led to misclassification of behaviors.
The community-based recruitment method of this study hopefully provided an accurate reflection of young urban IDUs. However, the potential for selection bias exists in this study because those who chose to participate in the DUIT trial may not be representative of all young drug users in Baltimore. This bias may, in part, explain the low STI and HIV seroprevalence observed in the current study. However, the utilization of a variety of recruitment techniques and the similarity of our seroprevalence data with other studies of young drug users support the representativeness of our cohort.
Despite these limitations, we documented a high prevalence of sexual risk behaviors among young IDUs in Baltimore. Our data suggest that HIV sexual risk reduction messages need to be enhanced to prevent acquisition of HIV and STIs among young drug users. In addition, further research on the importance and implications of douching-prevention messages among women at high risk for HIV infection is recommended.
1. Tyndall MW, Patrick D, Spittal P, et al. Risky sexual behaviours among injection drugs users with high HIV prevalence: implications for STD control. Sex Transm Infect 2002; 78(suppl 1):i170–175.
2. Latka M, Ahern J, Garfein RS, et al. Prevalence, incidence, and correlates of chlamydia and gonorrhea among young adult injection drug users. J Subst Abuse Treat 2001; 13:73–88.
3. Flom PL, Friedman SR, Kottiri BJ, et al. Stigmatized drug use, sexual partner concurrency, and other sex risk network and behavior characteristics of 18- to 24-year-old youth in a high-risk neighborhood. Sex Transm Dis 2001; 28:598–607.
4. Rottingen JA, Cameron DW, Garnett GP. A systematic review of the epidemiologic interactions between classic sexually transmitted diseases and HIV: How much really is known? Sex Transm Dis 2001; 28:579–597.
5. Dickerson MC, Johnston J, Delea TE, et al. The causal role for genital ulcer disease as a risk factor for transmission of human immunodeficiency virus. An application of the Bradford Hill criteria. Sex Transm Dis 1996; 23:429–440.
6. Martin HL, Richardson BA, Nyange PM, et al. Vaginal lactobacilli, microbial flora, and risk of human immunodeficiency virus type 1 and sexually transmitted disease acquisition. J Infect Dis 1999; 180:1863–1868.
7. Taha TE, Gray RH, Kumwenda NI, et al. HIV infection and disturbances of vaginal flora during pregnancy. J Acquir Immun Defic Syndr Hum Retrovirol 1999; 20:52–59.
8. Laga M, Manoka A, Kivuvu M, et al. Non-ulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: Results from a cohort study. AIDS 1993; 7:95–102.
9. Heckathorn DD. Respondent-driven sampling: A new approach to the study of hidden populations. Social Problems 1997; 44.
10. US Department of Health & Human Services. HIV Counseling, Testing, and Referral Standards and Guidelines, 1994.
11. Des Jarlais D, Paone D, Milliken J, et al. Audio-computer interviewing to measure risk behaviour for HIV among injecting drug users: A quasi-randomised trial. Lancet 1999; 353:1657–1661.
12. Metzger D, Koblin BA, Turner CF, et al. Randomized controlled trial of audio computer-assisted self-interviewing: Utility and acceptability in longitudinal studies. Am J Epidemiol 2000; 152:99–106.
13. Macalino GE, Celentano DD, Latkin C, et al. Risk behaviors by audio computer-assisted self-interviews among HIV-seropositive and HIV-seronegative injection drug users. AIDS Educ Prev 2002; 14:367–378.
14. Madico G, Quinn TC, Rompalo A, et al. Diagnosis of Trichomonas vaginalis
infection by PCR using vaginal swab samples. J Clin Microbiol 1998; 36:3205–3210.
15. Van Der Pol B, Ferrero DV, Buck-Barrington L, et al. Multicenter evaluation of the BDProbeTec ET System for detection of Chlamydia trachomatis
and Neisseria gonorrhoeae
in urine specimens, female endocervical swabs, and male urethral swabs. J Clin Microbiol 2001; 39:1008–1016.
16. Cosentino LA, Landers DV, Hillier SL. Detection of Chlamydia trachomatis
and Neisseria gonorrhoeae
by strand displacement amplification and relevance of the amplification control for use with vaginal swab specimens. J Clin Microbiol 2003; 41:3592–3596.
17. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol 1991; 29:297–301.
18. Schwebke JR, Hillier SL, Sobel JD, et al. Validity of the vaginal Gram stain for the diagnosis of bacterial vaginosis. Obstet Gynecol 1996; 88:573–576.
19. Joesoef M, Hillier SL, Josodiwondo S, et al. Reproducibility of scoring system for Gram stain diagnosis of bacterial vaginosis. J Clin Microbiol 1991; 29:1730–1731.
20. Schmid G, Markowitz L, Joesoef R, et al. Bacterial vaginosis and HIV infection. Sex Transm Infect 2000; 76:3–4.
21. Bachmann LH, Lewis I, Allen R, et al. Risk and prevalence of treatable sexually transmitted diseases at a Birmingham substance abuse treatment facility. Am J Public Health 2000; 90:1615–1618.
22. Hwang LY, Ross MW, Zack C, et al. Prevalence of sexually transmitted infections and associated risk factors among populations of drug abusers. Clin Infect Dis 2000; 31:920–926.
23. Lally MA, Alvarez S, Macnevin R, et al. Acceptability of sexually transmitted infection screening among women in short-term drug treatment. Sex Transm Dis 2002; 29:752–755.
24. Poulin C, Alary M, Bernier F, et al. Prevalence of Chlamydia trachomatis, Neisseria gonorrhoeae
, and HIV infection among drug users attending an STD/HIV prevention and needle-exchange program in Quebec City, Canada. Sex Transm Dis 1999; 26:410–420.
25. Hardick J, Hsieh YH, Tulloch S, et al. Surveillance of Chlamydia trachomatis
and Neisseria gonorrhoeae
infections in women in detention in Baltimore, Maryland. Sex Transm Dis 2003; 30:64–70.
26. Tracking the Hidden Epidemics: Trends in STDs in the United States.
Atlanta: Division of STD, Centers for Disease Control and Prevention, April 2001.
27. Wendel KA, Erbelding EJ, Gaydos CA, et al. Use of urine polymerase chain reaction to define the prevalence and clinical presentation of Trichomonas vaginalis
in men attending an STD clinic. Sex Transm Infect 2003; 79:151–153.
28. Joyner JL, Douglas JM Jr, Ragsdale S, et al. Comparative prevalence of infection with Trichomonas vaginalis
among men attending a sexually transmitted diseases clinic. Sex Transm Dis 2000; 27:236–240.
29. Turner CF, Rogers SM, Miller HG, et al. Untreated gonococcal and chlamydial infection in a probability sample of adults. JAMA 2002; 287:726–733.
30. Miller WC, Ford CA, Morris M, et al. Prevalence of chlamydial and gonococcal infections among young adults in the United States. JAMA 2004; 291:2229–2236.
31. Krieger JN, Alderete JF. Trichomonas vaginalis
and trichomoniasis. In: Holmes KK, Sparling PF, Mardh PA, et al., eds. Sexually Transmitted Diseases, 3rd ed. New York: McGraw-Hill, 1999:587–604.
32. Zhang J, Thomas AG, Leybovich E. Vaginal douching and adverse health effects: A meta-analysis. Am J Public Health 1997; 87:1207–1211.
33. Martino JL, Vermund SH. Vaginal douching: evidence for risks or benefits to women’s health. Epidemiol Rev 2002; 24:109–124.
34. Ness RB, Hillier SL, Richter HE, et al. Douching in relation to bacterial vaginosis, lactobacilli, and facultative bacteria in the vagina. Obstet Gynecol 2002; 100:765.
35. Holzman C, Leventhal JM, Qiu H, et al. Factors linked to bacterial vaginosis in nonpregnant women. Am J Public Health 2001; 91:1664–1670.
36. Schwebke JR, Desmond RA, Oh MK. Predictors of bacterial vaginosis in adolescent women who douche. Sex Transm Dis 2004; 31:433–436.
37. Hawes SE, Hillier SL, Benedetti J, et al. Hydrogen peroxide-producing lactobacilli and acquisition of vaginal infections. J Infect Dis 1996; 174:1058–1063.
38. Abma J, Chandra A, Mosher W, et al. Fertility, family planning, and women’s health: New data from the 1995 National Survey of Family Growth. Vital Health Stat 1997; 23:1–114.
39. Aral SO, Holmes KK. Social and behavioral determinants of the epidemiology of STDs: Industrialized and developing countries. In: Wasserheit JN, ed. Sexually Transmitted Diseases, 3rd ed. New York: McGraw-Hill, 1999.