Peterman, Thomas A. MD, MSc*; Newman, Daniel R. MA*; Goldberg, Martin BS†; Anschuetz, Greta L. MPH†; Salmon, Melinda BA†; Satterwhite, Catherine L. MPH, MSPH*; Berman, Stuart M. MD, ScM*
Chlamydia trachomatis is the most commonly reported notifiable infection in the United States.1 Infection with Chlamydia trachomatis is often asymptomatic but may lead to pelvic inflammatory disease (PID), ectopic pregnancy, and infertility.2 Screening high-risk women for C. trachomatis has been associated with decreased PID, and annual screening is recommended for all sexually active women under age 26.2–6 However, the impact of this screening at the population level is not so clear.7 PID has decreased since 1985, but there are multiple causes of PID and the diagnosis is nonspecific.8 The number of C. trachomatis infections reported continues to rise, but screening detects infections that might otherwise go undetected, so increased screening can increase the number of reported cases, even while the prevalence decreases.
C. trachomatis prevalence has been monitored in sentinel populations, such as women who are screened in family planning clinics.1 Test positivity among women attending family planning clinics fell when screening was introduced in the Pacific Northwest and in Columbus, OH, in the 1980s.9,10 However, test positivity has since remained flat in most regions.11 Some researchers suggest that screening women is not enough, and programs should also screen men.10,12
Screening men for C. trachomatis is difficult because men infrequently seek health care. However, some men do go to prison, and prisons have been good places to find infections.13,14 In New York City, a prison screening program for men less than 36 years of age identified 3854 infections in 2005, increasing the city-wide number of reported cases among men by 59%.15 However, each year the prison screening program tested approximately 25,000 men (of all ages), it found over one-third of all reported C. trachomatis infections among 20-24 year old men, and it detected C. trachomatis in 1.5% of all 20-24 year old men living in the high treatment ZIP codes. The objective of this study was to determine if this screening program influenced the likelihood of chlamydia infection among women from their neighborhoods who were tested in family planning clinics.
We divided the city into areas based on the percentage of men who were diagnosed and treated for C. trachomatis via the prison screening program. We then evaluated the changes in test positivity among young women from those areas who were tested for C. trachomatis in family planning clinics. Our retrospective analysis used data collected by the Philadelphia STD control program and Region III of the Infertility Prevention Project (IPP). We analyzed C. trachomatis test results from 20- to 24-year-old women tested in family planning clinics operated by IPP to see if test positivity decreased after the introduction of the prison screening program. Our plan of evaluation was developed, discussed, refined, and presented to others before data analysis.
The population of Philadelphia was 1,517,550 in 2000.16 The Philadelphia prison system released an average of 30,000 prisoners per year between 1996 and 2003.17 These prisoners were: men (86%); 18 to 24 years old (30%); black (70%), white (20%), and Hispanic (10%).17 The prison functions like jails in other jurisdictions; many prisoners stay only a short time. The most common reason for incarceration in 2003 was “pretrial hold” (n = 19,098), for whom the median days served was 17. Prisoners who were released before treatment were sought by the STD control program; 91.0% of 20- to 24-year-old men with identified infections were treated. Infected persons were given either a single dose of azithromycin (1 g) or 14 doses of doxycycline (100 mg) to be taken twice per day for a week.
The prevalence of C. trachomatis among 20- to 24-year-old women was estimated using the test positivity for C. trachomatis among young women tested in family planning clinics.18 Annual C. trachomatis screening has been recommended since 1993 for all sexually active women less than 20 years of age and for 20- to 24-year-old women who reported inconsistent use of barrier contraception or more than one sex partner in the last 3 months.19 This recommendation was expanded to all sexually active women under age 26 in 2002.20 Family planning clinics in Philadelphia have received federal funds from the IPP to screen for C. trachomatis since 1994, and since 1995, the policy in Philadelphia had been to screen all women under age 30 who attend family planning clinics and all women under age 35 who attend STD clinics. Starting in 1997, at least 98% of all tests done each year in family planning clinics were nucleic acid amplification tests. Other C. trachomatis screening programs implemented and maintained in Philadelphia after 1994 include screening in: the juvenile detention facility (2000); adult prisons (2002); public schools (2003); and family court (2004). In the first year, screening in Philadelphia public high schools led to the diagnosis and treatment of 766 young women and 246 young men.21 We excluded 15- to 19-year-old women from our analysis because we were concerned that the other screening programs, especially the high school program, might detect infections among young women that would have otherwise been detected in the family planning clinics, and thus give the appearance of a decline in community prevalence. This same phenomenon could also occur for the prison screening program, but 86% of prisoners were men.17
We defined geographic areas using ZIP codes of residence for men found to be infected in prison. The 2000 census provided estimates of the number of men and women of different ages living in those ZIP codes.16 For each ZIP code, we estimated the proportion of men treated for Chlamydia in prison by dividing the number of infections treated among 20- to 24-year-old men in prison by the number of men of that age living in that ZIP code. The 47 ZIP codes were ranked based on the proportions of men living in each area that were treated in prison (without regard to population size or geographic location). The highest 18 (“high-treatment”) were compared to the lowest 18 (“low-treatment”) ZIP codes. Data from the middle 11 ZIP codes were used to develop the analytic approach and were excluded from the final analysis.
Family planning clinics began screening at low levels in the 1990s, increased to over 26,000 tests per year in 1996, and increased again to over 36,000 tests per year in 1999. The number of 20- to 24-year-old women tested, and the type of test technology used was fairly consistent from 1999 through 2004. Thus, we used C. trachomatis test positivity from 1999 to 2001 to project the positivity we would expect in 2002 to 2004 without any intervention. Areas where many infections were treated due to the prison screening program were compared with the areas where few infections were treated in prison because we expected that the intervention would have the greatest impact in the high-screening areas. Trends that equally affected both areas would more likely be due to factors other than the prison screening program.
Among men of all ages, the numbers of tests (per year) done in Philadelphia prisons were: 16,860 (2002); 27,721 (2003); and 25,700 (2004). ZIP code information was generally not available for men who tested negative, but was available for all men who tested positive. From 2002 to 2004, there were 4263 positive C. trachomatis tests reported for men aged 20 to 24 years old living in Philadelphia; 1479 (34.7%) were diagnosed in prison and 1346 (91.0%) of them were treated. There were 23,203 men aged 20 to 24 years living in the 18 high-treatment ZIP codes and 1054 prison-identified infections were treated among men from those areas between 2002 and 2004. Thus, each year 1.5% of all 20- to 24-year-old men living in the high-treatment areas were treated for C. trachomatis that was diagnosed in prison (range among areas 0.92%-2.9%). (Table 1) There were 21,057 men aged 20 to 24 years living in the 18 low-treatment ZIP codes, and 98 prison-identified infections were treated among men from those areas between 2002 and 2004. Thus, each year 0.12% of all 20- to 24-year-old men living in those areas were treated for C. trachomatis that was diagnosed in prison (range among areas, 0%-0.28%). The 20- to 24-year-old men living in the high-treatment areas were 12.5 times more likely to have C. trachomatis detected in prison and treated than men living in the low-treatment areas. The high-treatment ZIP codes were clustered in 2 groups and the low-treatment ZIP codes were clustered in 3 groups. (Fig. 1).
From 1999 to 2004 there were 16,764 positive C. trachomatis tests reported for women aged 20 to 24 years old living in Philadelphia; 5664 (33.8%) were diagnosed in family planning clinics. During that time, family planning clinics reported 70,309 C. trachomatis tests for 20- to 24-year-old women, representing an average of 0.19 tests per year, for each of the 62,657 women of those ages living in Philadelphia (though some women might have been tested more than once in a year). Testing among 20- to 24-year-old women in family planning clinics remained fairly stable from 1999 to 2004, ranging from 11,056 to 12,119 tests per year (Table 1). During that time period, 56 clinics had done tests for at least one 20- to 24-year-old woman. Over half of all tests among 20- to 24-year-old women (53%) were done at 8 clinics. Nearly all clinics had tested women from both high- and low-treatment areas. The number of women from each ZIP code that were tested in family planning clinics remained fairly stable from 1999 to 2004.
Among 20- to 24-year-old women in the 18 high-treatment areas and 18 low-treatment areas, there were no major changes in the annual number of tests done in family planning clinics between 1999 and 2004 (Table 1), and there was no apparent shift in testing from one ZIP code to another (data not shown).
Test positivity for women in the high-treatment areas declined over time from 10.6% in 1999, to 8.7% in 2001, and 7.4% in 2004. The relative rate of annual decrease was greater before the prison screening intervention (18.2% or an average of 9.1% per year from 1999 to 2001) than during the intervention (14.6% or 4.9% per year from 2001 to 2004) (Fig. 2).
Test positivity for women in the low-treatment areas declined over time from 7.3% in 1999, to 5.4% in 2001, and 4.2% in 2004. Again, the relative rate of annual decrease was greater before the prison screening intervention (26.4% or 13.2% per year from 1999 to 2001) than during the intervention (22.6% or 7.5% per year from 2001 to 2004) (Fig. 2).
Although 20- to 24-year-old men living in the high-treatment areas were 12.5 times more likely to have C. trachomatis treated in prison than men living in the low-treatment areas, some of that difference in treatment might be due to the differences in the prevalence of infection among men in those areas. We do not know the prevalence in those areas and do not have ZIP codes for prisoners with negative tests. However, if we assume that the prevalence in men is proportional to the prevalence in women, we can adjust for the difference in prevalence among women tested in family planning clinics. From 2002 through 2004, women in high-treatment areas were 1.6 times more likely to test positive than women from low-treatment areas (Table 1). Adjusting for this suggests that infected men living in the high-treatment areas were 12.5/1.6 = 7.8 times more likely to be treated for C. trachomatis than infected men living in the low-treatment areas.
C. trachomatis test positivity among 20- to 24-year-old women tested in family planning clinics continued to decrease after men living in the same area were treated for C. trachomatis through the prison screening program; however, we found no evidence that the decrease was due to the new screening program. Test positivity decreased at a faster rate before the prison screening began than it did after the program was underway. Also, there were similar changes in test positivity among women from areas where a high percentage, compared with a low percentage, of men were treated for C. trachomatis due to the screening program. There are many possible reasons for this lack of change that fall into 2 broad categories: either the methods were inadequate to detect a true change or the intervention really did not make a significant difference.
This was an observational study, so many other factors could influence test positivity among women attending family planning clinics, including changes in test sensitivity or specificity, clinic screening practices, or attendance patterns in the clinics. We do not have any evidence that any of these occurred in a systematic way. Our study design assumed that, without the intervention, expected positivity would be predicted by trends before the intervention, and that the effects of prison screening would be greatest in areas where the most men were treated. We also assumed that unmeasured changes in testing practices would have a similar impact on high-treatment and low-treatment areas. Those assumptions may be wrong. It is possible that, without prison screening, test positivity would have increased in the high-treatment areas in comparison to the past, and in comparison to low-treatment areas. Our approach involves considerable uncertainty in estimating the expected test positivity, resulting in low statistical power. Regional data collection methods changed in 2005, and we could not reliably control for those changes, so we could not include data from after 2004.
There are several possible reasons why the intervention might not make a significant difference. There might be little overlap between the sexual networks of prisoners and the sexual networks of women who attend family planning clinics. Perhaps the men living in the high-treatment ZIP codes did not preferentially have sex with women from high-treatment ZIP codes. We do not know where men from these neighborhoods met their partners, but a study of persons with syphilis found they traveled a median of 1 mile to meet their partners.22 The 20- to 24-year-old men may have had sex with younger women who were not included in the study, however, older men were also screened in prison and had nearly identical high- and low-treatment ZIP codes. It is possible that many of the men acquired a new infection soon after treatment,23 and screening combined with expedited partner therapy24 might have a greater impact. Perhaps, the testing program was too small to expect a difference and the addition of screening at colleges and other places where men can be reached might have a bigger impact. However, each year the prison screening program tested approximately 25,000 men (of all ages) and found over one-third of all reported C. trachomatis infections among 20- to 24-year-old men and it detected C. trachomatis in 1.5% of all 20- to 24-year-old men living in the high-treatment ZIP codes. So, it was a big intervention.
Screening in prisons is a good way to find men infected with C. trachomatis; however, the impact on the prevalence of infection among women living in the community is more difficult to determine. We are aware of only one other study that addressed this issue. A jail-based screening program in San Francisco found 239 infections per year among 18-30 year-old men incarcerated between 1997 and 2004.25 During that time, the Chlamydia test positivity decreased from 16.1% to 7.8% among women attending a family planning clinic in the neighborhood where most of the prisoners lived while it remained stable at another clinic.
An ideal study design to assess the impact of screening in prisons would be a trial with random assignment of prisons to screen men for C. trachomatis or serve as controls. The prevalence of infection among women living in those communities would be measured before and after the male screening program. If screening was randomly assigned, and the only change during the trial was the addition of screening in select areas, then significant changes in prevalence in the intervention areas compared to the control areas could be attributed to the screening. Until such a study is done, we can only estimate effects using observational studies. We hope that future observational studies can address some of the limitations encountered in our analysis, but finding solutions to these problems will not be easy.
From 1999 to 2004, C. trachomatis test positivity for women attending family planning clinics decreased in Philadelphia while it remained stable in other areas.11 We found no evidence that the decrease was due to the prison screening program that began in 2002. Further research is needed to improve methods of measuring the impact of screening and to determine whether prison screening might make a difference in communities that are not experiencing a decrease in test positivity. In addition, research is needed to determine what was the cause of that major decrease in test positivity in Philadelphia.
1. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance, 2006. Atlanta, GA: US Department of Health and Human Services, 2007.
2. Peipert JF. Genital chlamydial infections. N Engl J Med 2003; 349:2424–2430.
3. Scholes D, Stergachis A, Heidrich FE, et al. Prevention of pelvic inflammatory disease by screening for cervical chlamydial infection. N Engl J Med 1996; 334:1262–1366.
4. Ostergaard L, Andersen B, Moller JK, et al. Home sampling versus conventional swab sampling for screening of Chlamydia trachomatis in women: A cluster-randomized 1-year follow-up study. Clin Infect Dis 2000; 31:951–957.
5. Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines, 2006. MMWR 2006; 55(No. RR-11):38.
6. Meyers DS, Halvorson H, Luckhaupt S. Screening for chlamydial infection: An evidence update for the US Preventive Services Task Force. Ann Intern Med 2007; 147:135–142.
7. Low N. Screening programmmes for chlamydial infection: When will we ever learn? BMJ 2007; 334:725–728.
8. Sutton MY, Sternberg M, Zaidi A, et al. Trends in pelvic inflammatory disease hospital discharges and ambulatory visits, United States, 1985-2001. Sex Transm Dis 2005; 32:778–784.
9. Britton TF, DeLisle S, Fine D. STDs and family planning clinics: A regional program for Chlamydia control that works. Am J Gynecol Health 1992; 6:80–87.
10. Mertz KJ, Levine WC, Mosure DJ, et al. Trends in the prevalence of chlamydial infections: The impact of community-wide testing. Sex Transm Dis 1997; 24:169–175.
11. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance, 2004. Atlanta, GA: US Department of Health and Human Services, 2005:13.
12. Miller WC. Screening for chlamydial infection: Are we doing enough? Lancet 2005; 365:456–458.
13. Bernstein KT, Chow JM, Ruiz J, et al. Chlamydia trachomatis and Neisseria gonorrhoeae infections among men and women entering California prisons. Am J Public Health 2006; 96:1862–1866.
14. Mertz KJ, Voigt RA, Hutchins K, et al. Findings from STD screening of adolescents and adults entering corrections facilities: Implications for STD control strategies. Sex Transm Dis 2002; 29:834–839.
15. Pathela P, Hennessy RR, Blank S, et al. The contribution of a urine-based jail screening program to citywide male Chlamydia and gonorrhea case rates in New York City. Sex Transm Dis 2009; 36:S58–S61.
17. Roman J, Kane M, Tukrner E, et al. Instituting lasting reforms for prisoner reentry in Philadelphia. Research Report Urban Institute, Justice Policy Center, Washington, DC, 2006.
18. Dicker LW, Mosure DJ, Levine WC. Chlamydia positivity versus prevalence: What’s the difference? Sex Transm Dis 1998; 25:251–253.
19. Centers for Disease Control and Prevention. Recommendations for the prevention and management of Chlamydia trachomatis infections, 1993. MMWR 1993; 42(No. RR-12):7.
20. Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines 2002. MMWR 2002; 51(No.RR-6):32.
21. Asbel LE, New Bern EC, Salmon M, et al. School-based screening for Chlamydia trachomatis and Neisseria gonorrhoeae among Philadelphia public high school students. Sex Transm Dis 2006; 33:614–620.
22. Michaud JM, Johnson SM, Ellen J. Comparison of sex partner meeting venues and residences of syphilis cases in Baltimore. Sex Transm Dis 2004; 31:239–242.
23. Peterman TA, Tian LH, Metcalf CA, et al. High incidence of new sexually transmitted infections in the year following a sexually transmitted infection: A case for rescreening. Ann Intern Med 2006; 145:564–572.
24. Golden MR, Whittington WLH, Handsfield HH, et al. Effect of expedited treatment of sex partners on recurrent or persistent gonorrhea or chlamydial infection. N Engl J Med 2005; 352:676–685.
25. Barry PM, Kent CK, Scott KC, Goldenson J, Klausner JD. Is jail screening associated with a decrease in chlamydia positivity among females seeking health services at community clinics?—San Francisco, 1997-2004. Sex Transm Dis 2009; 36:S22–S28.