IN THE UNITED STATES, an estimated three million persons acquire genital Chlamydia trachomatis infection annually. 1 In women, lower genital tract infection with C trachomatis frequently ascends to cause pelvic inflammatory disease and is a precursor to tubal infertility, ectopic pregnancy, and chronic pelvic pain. 2–4 Animal models and epidemiologic studies suggest that sequelae of genital chlamydial infection are more closely linked to second or subsequent infections than to initial infection. 5–7 Studies have described rates of persistent or recurrent infection of 5% to 38% in adolescents and young women treated for chlamydial infection. 8–13 However, data for these studies often were gathered retrospectively, follow-up periods were variable, incomplete behavioral data were available, and persistent infection due to therapeutic failure usually could not be distinguished from reinfection. The current study was designed to prospectively determine the frequency of persistence and reinfection with C trachomatis to identify preventable risk factors that might aid in the design of public health prevention strategies.
The study was undertaken at reproductive health, sexually transmitted disease (STD), and adolescent medicine clinics in Birmingham, Alabama; Indianapolis, Indiana; New Orleans, Louisiana; San Francisco, California; and Seattle, Washington. At each site, females 14 to 34 years were invited to participate if they had uncomplicated, laboratory-documented infection with C trachomatis. The method used to diagnose the initial chlamydial infection varied among cities and clinics; in most cases, infection was documented by culture, nucleic acid probe test (Pace II, GenProbe Corp., San Diego, CA), ligase chain reaction (LCR) (LCx, Abbott Laboratories, Chicago, IL) or polymerase chain reaction (Amplicor, Roche Laboratories, Nutley, NJ). Each subject provided informed consent, and the study was approved by the institutional review boards for human subjects research of each research center and that of the Centers for Disease Control and Prevention.
A prospectively recruited cohort design was used. Limited demographic, clinical, and behavioral data were collected from clinical and laboratory forms completed when the initial diagnosis of chlamydial infection was made, before patients were enrolled in the study. The study was introduced and patients enrolled when they were informed of their test results, usually 5 to 10 days after the screening visit, at which time treatment was given if presumptive treatment had not been administered at the screening visit. Patients were treated with standard regimens (usually, doxycycline 100 mg twice daily by mouth for 7 days or single-dose azithromycin 1.0 g orally) and were asked to return for the first study visit at 1 month after treatment and for a second study visit 4 months after the completion of treatment. Scheduled intervals for the two follow-up visits were 21 to 42 days and 2.5 to 5.5 months, respectively. Standardized demographic, clinical, and behavioral data were collected by interview at each study visit, including history of compliance with treatment, vaginal intercourse since treatment or since the preceding visit, condom use, new and continuing sexual partnerships, selected partnership characteristics, and patients’ beliefs about whether their partners had been treated for chlamydial infection. All centers used the same data collection forms and the interviewers were trained in its use. At all centers except San Francisco, women who returned were financially compensated (USD $10.00–$20.00 per visit) for their time and inconvenience. Women who failed to return for follow-up visits were contacted by mail, telephone, or in person to reschedule return visits. At each return visit, the first 30 ml of voided urine was collected for a C trachomatis LCR test, which was performed according to the manufacturer’s instructions. 14,15 Women with positive LCR test results at the first follow-up visit were considered to have completed the study and were not followed up further.
Data Management and Statistical Analysis
All data were entered into a single computer database (Epi Info version 6.04, Centers for Disease Control and Prevention and World Health Organization, Atlanta, GA). Dichotomous variables were compared by chi-square or Fisher’s exact test. Parametric or nonparametric tests were used as appropriate for the analysis of continuous variables. Multivariate techniques, including logistic regression, were used to investigate the relationship of factors while controlling for the effects of covariates. Mantel-Haenszel stratified techniques were used to estimate risk ratios and 95% CI. Attributable risk fractions were calculated using standard methods. 16
Study Population Characteristics
A total of 1,194 young women with documented C trachomatis infection agreed to participate in the study (Table 1). Recruitment by city ranged from 114 subjects in Indianapolis to 453 subjects in New Orleans. Those enrolled were most frequently seen at reproductive health clinics, most were black, and approximately half were age 19 years or younger.
First Study Visit
Among the 1,194 women enrolled, 622 (52.1%) returned for the first follow-up visit as scheduled (21–42 days after treatment). An additional 170 women (14.2%) returned after 42 days. Thus, 792 women (66.3%) completed the first return visit (mean follow-up interval, 38 days; range, 21–210 days). The first-visit return rates at the five study centers ranged from 40% (Birmingham) to 78% (New Orleans) (P < 0.001). Follow-up rates were lower for women initially seen at STD clinics (189 of 366 patients [51.6%]) than those seen at other facilities (610 of 828 patients [73.7%], P < 0.001). Additionally, women who acknowledged having two or more sex partners in the 60 days before enrollment were somewhat less likely to return (90 of 154 women [58.4%]) than those with one partner or no partner (696 of 1,027 women [67.8%], P = 0.03).
Chlamydial infection was documented by urine LCR in 50 of the 792 women (6.3%) who returned for the first follow-up visit. Table 2 shows the associations of selected demographic and behavioral variables with persistent or recurrent infection at that visit. Infection was identified in 7.8% of those who resumed sexual activity since the initial visit, compared with 3.7% of those who reported no sexual activity in the interim (unadjusted relative risk [RR], 2.1; 95% CI, 1.1–4.2). Infection at the first return visit also was significantly associated in univariate analysis with acknowledgment by the woman that she failed to complete therapy as instructed (RR, 3.2; 95% CI, 1.5–6.9), an association that remained after controlling for treatment regimen. Women age 19 years or younger were slightly more likely to be have a chlamydial infection than women 20 years or older (RR, 1.3; 95% CI, 0.8–2.3), as were women who returned after the intended appointment window (RR, 1.7; 95% CI, 0.9–3.0). A self-report of having talked to male sex partners about the partners’ need for treatment, being told by partners that they were treated, and condom use during the most recent vaginal intercourse were not significantly associated with infection, nor was the age differential between the women and their partners (Table 2).
In multivariate analysis that included age, completion of treatment, resumption of sexual activity, and return outside the intended appointment window, independent predictors of chlamydial infection at the first return visit were failure to complete treatment (RR, 3.4; 95% CI, 1.6–7.3), resuming sexual activity since completing treatment (RR, 2.0; 95% CI, 1.03–4.4), and return beyond the intended appointment window (RR, 1.8; 95% CI, 1.01–3.2). Those 19 years and younger were at somewhat greater risk of infection (RR, 1.4; 95% CI, 0.8–2.4) than women 20 years and older. Other behavioral factors, the type of clinical facility where the diagnosis was made, and research center did not significantly affect estimates of risk. Among the women who resumed sexual activity, 7.8 per 100 women were infected, compared with 3.7 per 100 women who did not report that they resumed sexual activity. Thus, the risk difference was 4.1 (95% CI, 0.9–7.4) and the univariate population attributable fraction due to the resumption of sex was 42% (2.7 infections per 100 women). The remaining infections likely were due to treatment failure.
Second Study Visit
Of 742 women with negative test results for chlamydial infection at the first return visit, 419 (56.5%) were seen again within the intended interval of 2.5 to 5.5 months after treatment. An additional 86 women were seen before or after the intended interval. Thus, 505 women (68.1%) had a second visit 4 weeks or longer after the first return visit (median follow-up interval, 130 days [4.3 months] after initial infection; range, 70–408 days). An additional 21 women who returned but gave histories of repeat treatment for chlamydial infection after the first return visit; all tested negative for chlamydial infection at the second return visit and were excluded from analysis. The proportion of patients who returned for the second follow-up visit ranged from 42.8% to 72.5% among the five research centers. The 505 evaluable women who returned for the second return visit did not differ significantly in any demographic or behavioral variables from all enrolled subjects (data not shown).
Chlamydial infection was diagnosed by urine LCR at the second visit in 36 (7.1%) of 505 women (Table 3). Mean follow-up intervals since the first return visit were 92 days and 94 days for women with and without infection, respectively. Risk factors for recurrent infection that approached statistical significance by univariate analysis were an age of 19 years or younger (RR, 1.7; 95% CI, 0.9–3.3) and a history of being sexually active with at least one partner between the two visits (RR, 2.2; 95% CI, 0.7–7.1). Younger women were slightly less likely to resume sexual activity; among 224 women 19 years and younger, 175 (78.1%) acknowledged having sex since the first follow-up visit, compared with 241 (87.3%) of 276 women 20 years or older (P < 0.01). Among women who were sexually active after the first return visit, 19 (10.9%) of 175 women 19 years or younger were reinfected, compared with 14 (5.8%) of 241 women 20 years and older (P = 0.07). Age of sex partners, number of episodes of vaginal intercourse reported, and frequency of condom use were similar between infected and uninfected women (data not shown).
In multivariate analysis that included age, resumption of sexual activity, and sex with new partners, being 19 years or younger was modestly associated with infection (RR, 1.8; 95% CI, 0.97–3.5;P = 0.08). Resumption of sexual activity was not significantly associated with infection (RR, 2.4; 95% CI, 0.8–7.7;P > 0.1). Having a new sex partner since the first return visit was not significantly associated with infection at the second visit (RR, 1.3; 95% CI, 0.7–2.6), and the univariate population attributable fraction due to a new sex partner was only 5.5%, suggesting that few infections could be attributed to this risk variable. Further, infection was not independently associated with having sex with partners that the woman classified as casual (nonsteady) or the use of a condom during the most recent episode of intercourse. At the first return visit, 426 of 493 women (86.4%) reported talking to all of their sex partners about the partners’ need for treatment, and 322 women (65.3%) reported that one or more of their partners said they had been treated. Neither the patients’ first return-visit report that they spoke with their partners about the partners’ need for treatment nor the patients’ belief that their partners were treated was associated in multivariate analysis with protection against reinfection at the second return visit.
This study documents a high rate of persistent or recurrent infection with C trachomatis in a cohort of young women followed up prospectively in five geographically diverse areas of the United States. Overall, 13.4% of women had persistent infection or reinfection within a median of 4.3 months of treatment for uncomplicated chlamydial infection. At the first posttreatment evaluation, which occurred at a median of 38 days after treatment, 6.3% of evaluated women were infected, a rate similar to that observed in a recent randomized clinical trial that measured outcome 1 month after treatment. 13 Infection at that visit was independently associated with incomplete medication compliance and with resumption of sexual activity, usually in a continuing sexual partnership, and teens were at somewhat higher risk than older women. These findings suggest that infection in the first few weeks after treatment resulted from reinfection both from untreated sex partners and from treatment failure. The frequency of persistent or recurrent infection was similar in patients treated with azithromycin, doxycycline, or other regimens. Different frequencies of chlamydial infection in women who had or had not resumed sexual activity permitted estimation of the fraction of infections due to therapeutic failure. The estimate of 3.7% is consistent with observations from contemporary therapy trials. 17,18 Therefore, some persistent infections probably are caused by failure to eradicate C trachomatis, despite compliance with treatment.
We considered all positive LCR test results for C trachomatis infection at the second return visit (median, 4.3 months after treatment and 3.0 months after a documented negative test at the first return visit) to be the result of reinfection, rather than persistent infection. These infections were not significantly associated with any of several partnership-linked variables. However, the frequencies of certain risk factors were lower in our study than in some other recent reports. 19,20 For example, only 8.6% of participants in this study reported multiple sex partners, and only 28.2% reported intercourse with a new partner during the interval between a negative LCR test result and the second study visit. Therefore even after several months, most C trachomatis reinfections in women in this study were acquired from the same male sex partner(s) they had when their infection was first diagnosed, an observation reported by other investigators. 8,21 This study was not designed to determine whether the partners had persistent (untreated) infections or had acquired new infections from other partners.
These observations suggest that the women’s sexual behaviors were not the primary determinants of reinfection, which implies that their partners’ sexual or healthcare behaviors were the more important determinants. Eighty-six percent of women reported speaking to their male partners about his need for treatment, and 65% reported that their partners said that they had been treated. However, we did not contact the partners, and this study was not designed to measure the actual sexual or healthcare-seeking behaviors of their male partners.
The interactions of women and men concerning STD treatment are complex; for example, fear of violence or fear of loss of the support of a relationship can inhibit a woman’s willingness to be assertive in asking her partner to seek treatment. Men and women respond differently to the risk message implicit in an STD diagnosis. 22 The availability of STD-linked healthcare services for men is limited in many settings. For example, lengthy waits at public STD clinics can be disincentives to receiving necessary care, and men may feel unwelcome or otherwise uncomfortable seeking care from their partners’ reproductive health providers. Further, urethral chlamydial infection in men often is asymptomatic, and the absence of symptoms may be a further disincentive to seeking care. Thus, further understanding of the steps needed to reduce the risk of reinfection in women will need to address the behaviors of both women’s sex partners and the women themselves. In any case, our results highlight the importance of assuring treatment of women’s male sex partners. Retrospective studies suggest that providing infected women with antibiotics to deliver to their male partner can reduce rates of reinfection in women, 23,24 and a prospective multicenter randomized trial of this strategy is underway.
The results of this study confirm that persistent and recurrent infection with C trachomatis is common among women who are treated for uncomplicated chlamydial infection. Approximately one in seven women remained infected or became reinfected at a rate equivalent to 33 infections per 1,000 person months of observation. Despite major differences in study design, this rate approximates the incidence estimate of 28 new infections per 1,000 person months that was recently reported by Burstein et al 11 in a longitudinal study of inner-city adolescents. We recruited and prospectively followed a cohort of young women through two follow-up visits, whereas Burstein et al retested a convenience sample of patients when they returned for any healthcare visit at selected facilities, without active follow-up efforts.
This study has several limitations, and further studies will be needed to confirm and refine our results. The low frequency of certain risk behaviors may have contributed to insufficient power to detect some potentially important determinants of chlamydial persistence or reinfection. Second, although follow-up rates at both the first and second return visits were good in view of the characteristics of populations at risk for STD, only about two thirds of eligible women returned for each follow-up visit. However, women returning for the second follow-up visit did not differ from enrolled women in a number of demographic and behavioral characteristics. Third, the variability of the actual intervals at which patients returned for study visits created an analytic challenge, although controlling for return before or after the nominal intervals did not significantly affect the results. Fourth, several kinds of tests were used to detect the patients’ original chlamydial infections. Because the specificities of all these tests are high, almost all patients were infected. However, some of the tests used are insensitive, and some infections undoubtedly were missed. Therefore, the results of this study may not apply to all infected women, perhaps especially those with low concentrations of C trachomatis in cervical secretions, who may be more likely to have chronic or subclinical infections. Finally, at most study centers the participants received financial incentives, and all patients received services that may not be available to other women with chlamydial infection, including counseling that emphasized the importance of preventing future infections. These factors might have influenced risk-taking behaviors and reduced the reinfection rate below that experienced by other women. 25 Thus, in some settings the actual rates of persistent or recurrent chlamydial infection may be higher than we observed.
Rescreening, or routine retesting of women several weeks or months after treatment, may be an important strategy in the control of chlamydial infections and their complications. Rescreening is distinct from test-of-cure, wherein patients are tested after a short interval to detect treatment failure; except in pregnant women, test-of-cure is not recommended for chlamydial infection treated with standard regimens. 26 Rescreening of women has been recommended as a gonorrhea control strategy, but has rarely been implemented on a wide scale because of poor compliance by persons at risk, 27 which in turn may be due in part to the need for a clinic visit and a vaginal speculum examination. However, nucleic acid amplification tests for C trachomatis on urine or on self-collected vaginal swabs or tampons 28–30 permit testing without a speculum examination, and even without a clinic visit if specimen transport can be arranged. Rescreening necessitates only a modest increment in the total number of C trachomatis tests performed, equivalent to the prevalence in the population tested less the proportion of patients who do not comply with rescreening. The performance and cost effectiveness of rescreening need careful analysis, but this and most other reported studies 8–13 have documented rates of persistent or recurrent chlamydial infection well above the minimum prevalences that make primary screening programs cost effective. 31–33
Because we were unable to identify definitive determinants of recurrent infection, we recommend that rescreening be offered to all women treated for chlamydial infection, with emphasis on those who are likely to resume sexual activity after treatment. The best interval for a single rescreening test cannot be precisely defined by the available studies, but our results suggest that a test 2.5 to 5.5 months after treatment may be appropriate. Based partly on the results of this study, routine rescreening with urine or vaginal-swab LCR testing 3 to 4 months after treatment recently was implemented for all women with chlamydial infection at public health clinics in King County, Washington and San Francisco, California and is used at some of the current investigators’ own clinics. Although assuring compliance will be challenging and rescreening must not supplant continued primary screening of young women at risk, this approach should be adopted as a routine strategy for chlamydia prevention.
1. Groseclose SL, Saidi AA, DeLisle SJ, Levine WC, St. Louis ME. Estimated incidence and prevalence of genital Chlamydia trachomatis
infections in the United States, 1996. Sex Transm Dis 1999; 26: 339–344.
2. Seilors JW, Mahony JB, Chernesky MA, Rath DJ. Tubal factor infertility. Fertil Steril 1988; 49: 451–457.
3. Brunham RC, Maclean IW, Binns B, Peeling RW. Chlamydia trachomatis: its role in tubal infertility. J Infect Dis 1985; 152: 1275–1282.
4. Weström L, Mårdh PA. Acute pelvic inflammatory disease. In: Holmes KK, Mårdh PA, Sparling PF, et al, eds. Sexually Transmitted Diseases, 2nd ed. New York: McGraw Hill Book Co, 1990: 593–613.
5. Van Voorhis WC, Barrett LK, Sweeney YT, Kuo CC, Patton DL. Repeated Chlamydia trachomatis
infection of Macaca nemestrina
fallopian tubes produces a Th1-like cytokine response associated with fibrosis and scarring. Infect Immun 1997; 65: 2175–2182.
6. Hillis SD, Owens LM, Marchbanks PA, Amsterdam LF, MacKenzie WR. Recurrent chlamydial infections increase the risks of hospitalization for ectopic pregnancy and pelvic inflammatory disease. Am J Obstet Gynecol 1997; 176: 103–107.
7. Egger M, Low N, Smith GD, Lindblom B, Herrmann B. Screening for chlamydial infections and the risk of ectopic pregnancy in a county in Sweden: an ecological analysis. BMJ 1998; 316: 1776–1780.
8. Blythe MJ, Katz BP, Batteiger BE, Ganser JA, Jones RB. Recurrent genitourinary chlamydial infections in sexually active female adolescents. J Pediatr 1992; 121: 487–493.
9. Hillis SD, Nakashima A, Marchbanks PA, Addis DG, Davis JP. Risk factors for recurrent Chlamydia trachomatis
infection in women. Am J Obstet Gynecol 1994; 170: 801–806.
10. Orr DP, Langefeld CD, Katz BP, Caine VA. Behavioral intervention to increase condom use among high-risk female adolescents. J Pediatr 1996; 128: 288–295.
11. Burstein GR, Gaydos CA, Diener-West M, Howell MR, Zenilman JM, Quinn TC. Incident Chlamydia trachomatis
infections among inner-city adolescent females. JAMA 1998; 280: 521–526.
12. Oh MK, Cloud GA, Fleenor M, Sturdevant MS, Nesmith JD, Feinstein RA. Risk for gonococcal and chlamydial cervicitis in adolescent females: incidence and recurrence in a prospective cohort study. J Adolesc Health 1996; 18: 270–275.
13. Hillis SD, Coles FB, Litchfield B, et al. Doxycycline and azithromycin for prevention of chlamydial persistence or recurrence one month after treatment in women: a use-effectiveness study in public health settings. Sex Transm Dis 1998; 25: 5–11.
14. Schachter J, Moncada J, Whidden R, et al. Noninvasive tests for diagnosis of Chlamydia trachomatis
infection: application of ligase chain reaction to first-catch urine specimens of women. J Infect Dis 1995; 172: 1411–1414.
15. Lee HH, Chernesky MA, Schachter J, et al. Diagnosis of Chlamydia trachomatis
genitourinary infection in women by ligase chain reaction assay of urine. Lancet 1995; 345: 213–216.
16. Hennekens CH. Epidemiology in Medicine. New York: Little, Brown and Company, 1987: 77–96.
17. Martin DH, Mroczkowski TF, Dalu ZA, et al. A controlled trial of a single dose of azithromycin for the treatment of chlamydial urethritis and cervicitis. The Azithromycin for Chlamydial Infections Study Group. N Engl J Med 1992; 327: 921–925.
18. Thorpe EM Jr, Stamm WE, Hook EW III, et al. Chlamydial cervicitis and urethritis: single dose treatment compared with doxycycline for seven days in community based practices. Genitourin Med 1996; 72: 93–97.
19. Mosure DJ, Berman S, Kleinbaum D, Halloran J. Predictors of Chlamydia trachomatis
infection among female adolescents: a longitudinal analysis. Am J Epidemiol 1996; 144: 997–1003.
20. Marrazzo JM, Fine D, Celum CL, DeLisle S, Handsfield HH. Selective screening for chlamydial infection in women: a comparison of three sets of criteria. Fam Plann Perspect 1997; 29: 158–162.
21. Hook EW III, Reichert CA, Upchurch DM, Ray P, Celentano D, Quinn TC. Comparative behavioral epidemiology of gonococcal and chlamydial infections among patients attending a Baltimore, Maryland sexually transmitted disease clinic. Am J Epidemiol 1992; 136: 662–672.
22. Warszawski J, Meyer L. Gender difference in persistent at-risk sexual behavior after a diagnosed sexually transmitted disease. Sex Transm Dis 1998; 25: 437–442.
23. Kissinger P, Brown R, Reed K, et al. Effectiveness of patient delivered partner medication for preventing recurrent Chlamydia trachomatis.
Sex Transm Infect 1998; 74: 331–333.
24. Ramstedt K, Forssman L, Johannisson G. Contact tracing in the control of genital Chlamydia trachomatis
infection. Int J STD AIDS 1991; 2: 116–118.
25. Kamb ML, Fishbein M, Douglas JM Jr, et al. Efficacy of risk-reduction counseling to prevent human immunodeficiency virus and sexually transmitted diseases: a randomized controlled trial. JAMA 1998; 280: 1161–1167.
26. Centers for Disease Control, Prevention. guidelines for treatment of sexually transmitted diseases. MMWR Morb Mortal Wkly Rep 1998;1998: 47(RR-1): 53–59.
27. Wolfe F, Judson FN. Re-screening for gonorrhea: an evaluation of compliance methods and results. Am J Public Health 1979; 69: 1178–1180.
28. Hook EWIII, Smith K, Mullen C, et al. Diagnosis of genitourinary Chlamydia trachomatis
infections by using the ligase chain reaction on patient-obtained vaginal swabs. J Clin Microbiol 1997; 35: 2133–2135.
29. Stary A, Schuh E, Kerschbaumer M, Gotz B, Lee H. Performance of transcription-mediated amplification and ligase chain reaction assays for detection of chlamydial infection in urogenital samples obtained by invasive and noninvasive methods. J Clin Microbiol 1998; 36: 2666–2670.
30. Tabrizi SN, Paterson B, Fairley CK, Bowden FJ, Garland SM. A self-administered technique for the detection of sexually transmitted diseases in remote communities. J Infect Dis 1997; 176: 289–292.
31. Howell MR, Quinn TC, Brathwaite W, Gaydos CA. Screening women for Chlamydia trachomatis
in family planning clinics: the cost-effectiveness of DNA amplification assays. Sex Transm Dis 1998; 25: 108–117.
32. Paavonen J, Puolakkainen M, Paukku M, Sintonen H. Cost-benefit analysis of first-void urine Chlamydia trachomatis
screening program. Obstet Gynecol 1998; 92: 292–298.
33. Marrazzo JM, Celum CL, Hillis SD, Fine D, DeLisle S, Handsfield HH. Performance and cost-effectiveness of selective screening criteria for Chlamydia trachomatis
infection in women: implications for a national chlamydia control strategy. Sex Transm Dis 1997; 24: 131–141.