Chlamydia trachomatis and Neisseria gonorrhoeae Infections Among Men and Women Entering California Prisons

Abstract

Objective. We estimated the prevalence of Chlamydia trachomatis and Neisseria gonorrhoeae infection among newly arriving inmates at 6 California prisons.

Methods. In this cross-sectional study in 1999, urine specimens collected from 698 men aged 18 to 25 years and 572 women aged 18 years or older were tested at intake for C trachomatis and N gonorrhoeae using ligase chain reaction. An analysis of demographic and arrest-related correlates of C trachomatis and N gonorrhoeae infection was performed.

Results. The overall C trachomatis prevalence was 9.9% (95% CI=7.8%, 12.3%) among men aged 18 to 25 years, 8.9% (95% CI = 2.9%, 22.1%) among women aged 18 to 25 years, and 3.3% (95% CI=2.0%, 5.1%) among women overall. Three N gonorrhoeae cases were detected with an overall prevalence of 0.24% (95% CI=0.05%, 0.69%).

Conclusions. The prevalence of C trachomatis infection at entry to California prisons, especially among young female and male inmates, was high, which supports routine screening at entry into prison. In addition, screening in a jail setting where most detainees are incarcerated before entry into the prison setting may provide an excellent earlier opportunity to identify these infections and treat disease to prevent complications and burden of infection in this high-risk population.

In the United States in 2002, 1 of every 143 residents was in a state or federal prison or a local jail.1 The California Department of Corrections reported that the average monthly population in state prisons during December 2003 was approximately 160 000 with an 18-month average length of stay.2 In California, entry into any correctional setting begins with intake at a jail facility run by either city or county jurisdictions. After conviction for more serious offenses or convictions with sentences of greater than a year, detainees are then transferred to a state or federal prison.

Sexual risk behaviors such as multiple partners, history of sexually transmitted diseases (STDs), and lack of condom use before incarceration have been reported among males in correctional settings as well as in a national survey of inmates.3 In the only published study of risk behaviors among women inmates, more than half of the women entering a North Carolina state correctional facility reported at least 1 high-risk sexual behavior, including inconsistent condom use with multiple partners, history of STD, or exchanging sex for drugs or money.4

Cross-sectional studies of STD prevalence conducted mainly in large metropolitan county jail facilities have found high levels of Chlamydia trachomatis (C trachomatis) prevalence ranging from 7% to 27% and Neisseria gonorrhoeae (N gonorrhoeae) prevalence at approximately 5%.5^–7 Failure to detect STDs in the jail inmate population is of significant public health concern because this is an indication of missed opportunities to diagnose and treat this population before their return to the community and ultimately to prevent further transmission. Additionally, few studies of bacterial STD prevalence, as opposed to that of HIV and other viral STDs, have been conducted in prison settings.8 Older studies of bacterial STDs in prison populations found a N gonorrhoeae prevalence of 1.6% among males aged 25 years and younger9 and C trachomatis and N gonorrhoeae prevalence of 5% and 2%, respectively, among female inmates with a mean age of 24 years.10 These older findings probably underestimate the C trachomatis and N gonorrhoeae prevalence because of the use of less sensitive nonamplified diagnostic tests, and therefore suggest that C trachomatis and N gonorrhoeae prevalence may be significantly higher when more sensitive testing methods are used.

In spite of the high prevalence of STD risk behaviors and STDs among the incarcerated, STD screening services are inconsistently available in city and county jails. A 1997 Centers for Disease Control and Prevention survey of 94 US counties with high infectious syphilis morbidity or cities with populations of 200 000 or more, found that although a large percentage of jails had some STD screening policy, only between 0.2% and 0.6% of the jailed client population was actually screened.11 Jails were significantly more likely to offer syphilis screening to clients than N gonorrhoeae or C trachomatis screening. Less than 25% of facilities routinely offered C trachomatis or N gonorrhoeae screening to all clients, and among the facilities that reported routine STD screening, less than half of inmates were tested.

Despite national standards for STD screening issued by the National Commission on Correctional Health Care (NCCHC),12^–13 STD screening in correctional facilities is uncommon and well below these national standards. Possible explanations for low screening rates have included discharge of inmates before screening, and lack of staff, space, and testing resources to conduct STD screening, but few studies have rigorously examined the barriers to STD screening in corrections.11 Assessment of STD prevalence in inmates at entry to prison settings is needed to further support the earlier detection of these infections through jail STD-screening programs. We examined the association of demographic as well as population-level correlates such as C trachomatis rates in the county of arrest and arrest charges associated with exchanging sex for money or drugs; this information was readily available at prison entry and could be used for developing targeted screening strategies at entry.

We conducted a cross-sectional study to assess the prevalence of C trachomatis and N gonorrhoeae in prisons to guide screening recommendations and indirectly assess the success of screening at the local jail correctional setting. We used the more sensitive nucleic acid amplified tests and assessed demographic and arrest-related correlates of infection among newly arriving inmates in 2 women’s and 4 men’s prisons in California.

METHODS

Prisons

In California, site of entry into a correctional facility is determined by the county in which the crime was committed as well as by the nature of the crime. Local jails hold prisoners who are awaiting court processing, sentencing, or have been sentenced to less than 1 year. Adjudicated inmates who have been sentenced to greater than a year’s sentence move from the jail usually within a year to a federal prison or a state prison reception center housed within the prison facility. Here, they are assigned to a specific longer-stay prison on the basis of type of crime, level of education, violence level, health status, and other factors. Thirteen prison reception centers serve as entrance points into the California prison system.

The California Department of Health Services and the California Department of Corrections jointly selected the prison sites for this study. Six sites were selected by the following criteria: (1) representation of both urban and rural areas, (2) sufficient volume of newly arriving inmates, (3) representation of all 3 geographic regions (Northern [1 of 3 total], Central Valley [3 of 6 total], and Southern [2 of 4 total]) in California, (4) representation of men’s (4 of 10 total) and women’s (2 of 3 total) facilities, and (5) routine collection of urine specimens from all inmates during the physical examination performed at admission. Overall, the sites comprised 2 women’s prisons in Central Valley and Southern regions, and 4 men’s prisons in Northern, Central Valley (n=2), and Southern regions.

Study Population

Between January 25, 1999, and March 19, 1999, all inmates entering the 6 California prisons through the respective reception centers were eligible to participate in the study. This study was part of an ongoing, blinded sentinel surveillance project to monitor prevalence of HIV, hepatitis, and other STDs in California prisons. Verbal consent to perform blinded testing was obtained at the time of intake. The California Health and Human Services Agency Committee for the Protection of Human Subjects approved the study protocol.

Limited resources restricted the testing of all male entrants, so only men aged 25 years or younger entering 1 of the 6 prison sites were eligible to be tested. All women regardless of age were tested. Because daily census data for entering inmates were not available, we calculated the number of entering inmates based on the reported estimates that 97% of all inmates receive physical examinations, and 18.3% of male inmates are aged 25 years and younger (E. Horowitz, MD, oral communication, April 2005). Because of the blinded nature of the study, participants who may have been previously detained in the prison system during the study period could not be excluded. However, monthly residence data suggest that during a 3-month period this number would be quite low.

Clinical Evaluation

At the time of the study, the standard of care in all prisons was that all newly arrived prisoners received a brief assessment and physical examination shortly after arrival at a prison, which included a standard blood draw as well as urine sample collection. All inmates who were identified as infected through routine STD testing of these specimens were treated per standard medical protocol by corrections staff. (Because of the need to protect inmate confidentiality, these medical and laboratory records are not available for secondary analysis; California Health and Human Services Committee for the Protection of Human Subjects approval of the study protocol was based on testing of leftover specimens obtained at entry for routine medical evaluation.) During the enrollment period, incoming inmates were made aware that their leftover urine specimens were being saved for a blinded study of disease prevalence in California prisons and that laboratory test results could not be linked back to them by study analysts. Prison health care staff informed incoming inmates that they could refuse participation, in which case their leftover specimens would be destroyed.

Study Variables

At the time of physical examination, California Department of Corrections nursing staff abstracted demographic data and arrest history data from inmate medical records and prison databases that had been merged with laboratory test result data. The nurses also assigned unique study identifiers to the abstracted data. The demographic information collected included date of birth, gender, and race/ethnicity. Arrest history from California Department of Corrections databases included county of arrest and the penal code(s) for participants’ current arrest; however, date of arrest was not available. Because of the multiple penal codes recorded for current incarceration, 9 broad arrest categories based on the frequency of individual penal codes were created: assault, murder, theft, drug possession/sale, gun possession/sale, sex crimes, white collar crimes, driving while under the influence of alcohol, and “other” crimes. Records were assigned a unique study identifier and could not be linked back to the original inmate data.

To examine the relation between arrest in a county with high reported case rates of C trachomatis and C trachomatis infection at prison intake, a dichotomous variable was created: “arrest in a high CT county.” Because of the lack of C trachomatis screening guidelines to detect asymptomatic chlamydial infections in men, case-based surveillance data among male populations does not accurately reflect C trachomatis prevalence among men in California. Therefore, high C trachomatis counties were defined as counties whose 1999 C trachomatis rates were greater than the overall California state rate for women (249.9 per 100000).14

Laboratory Methods

Following the completion of the admission physical examination, leftover urine specimens were sent to the local health jurisdiction public health laboratory within 7 days of collection. There the specimens were aliquotted and either sent on a weekly basis (in the Northern region) to the Chlamydia Research Laboratory, University of California, San Francisco, for immediate testing, or stored in a −70 °C freezer (in the Central Valley and Southern regions) until transport for C trachomatis and N gonorrhoeae testing on a batch basis; these specimens remained frozen for approximately 2 to 3 weeks until testing. C trachomatis and N gonorrhoeae infections were detected by ligase chain reaction (LCx, Abbott Laboratories, Abbott Park, IL) according to previously published protocols.15

Statistical Analysis

All analyses were performed using SAS software version 8 (SAS Institute Inc, Cary, NC) or Stata software version 6 (Stata Corp, College Station, TX). The prevalence estimates of C trachomatis and N gonorrhoeae were calculated as the number of ligase chain reaction–positive participants divided by the number of participants whose urine samples were tested. Ninety-five percent confidence intervals (CIs) were generated from calculations for proportions derived from a binomial distribution. Crude and adjusted odds ratios (ORs) were used to estimate associations between disease status and correlates on the basis of logistic regression modeling and associated 95% CIs were calculated using Woolf’s formula.16 All analyses were stratified by gender.

RESULTS

During the study period, an estimated total of 5907 inmates were evaluated in the prison reception centers including 1080 men aged 25 years and younger and 775 women. After excluding 3% of inmates who were parole violators or combative, there were a total of 894 men aged 25 years and younger and 587 women who had physical examinations and urine specimens who were eligible for participation. There were 196 men and 15 women excluded because of refusals or specimen processing problems. The data analysis was performed on the remaining 1270 participants who had urine specimens tested for C trachomatis and N gonorrhoeae: 698 men (64.6%) and 572 women (73.8%). The majority of women (91.0 %) were aged older than 25 years (Table 1 ▶). Women were more likely to be African American, and less likely to be Hispanic, than were men (P < .001). Furthermore, men were more likely to have been arrested for assault, gun possession, murder, sex-related crimes, and theft than were women (P < .001). Women were more likely to be arrested for drug-related crimes than were men (P < .001).

TABLE 1—

Demographic and Arrest History Characteristics and Chlamydial Infection Among Inmates Entering 6 California Prisons, by Gender, 1999

	Men (n = 698)				Women (n = 572)
	No.	%CT positive	Unadjusted Odds Ratio	95% CI	No.a	%CT positive	Unadjusted Odds Ratio	95% CI
Age, years
≤25	698	9.9	. . .	. . .	45	8.9	3.3	0.9, 11.3
26–30	. . .	. . .	. . .	. . .	93	4.3	1.7	0.4, 5.9
> 30	. . .	. . .	. . .	. . .	427	2.6	1.0	Reference
Race/ethnicity*
White	87	5.8	1.0	Reference	183	2.2	1.0	Reference
African American	142	12.0	2.2	0.8, 6.3	196	4.1	1.9	0.6, 6.4
Hispanic	382	9.2	1.7	0.6, 4.4	144	2.8	1.3	0.3, 5.2
Asian/Pacific Islander	25	4.0	0.7	0.1, 6.1	2	0.0	. . .	. . .
American Indian	4	0.0	. . .	. . .	9	11.1	5.6	0.6, 56.0
Unknown	58	19.0	3.8	1.3, 11.7	38	5.3	2.5	0.4, 14.1
Arrest Chargeb
Assault*	122	8.2	0.8	0.4, 1.6	23	4.4	1.3	0.2, 10.5
DUI	13	0.0	. . .	. . .	6	0.0	. . .	. . .
Drug-related*	232	10.3	1.1	0.6, 1.8	344	3.5	1.1	0.4, 2.9
Gun possession*	64	7.8	0.8	0.3, 2.0	10	0.0	. . .	. . .
Homicide*	41	4.9	0.5	0.1, 1.9	8	0.0	. . .	. . .
Sex-related*	24	0.0	. . .	. . .	2	0.0	. . .	. . .
Theft*	366	12.0	1.7	1.0, 2.8	239	3.4	1.0	0.4, 2.6
White collar*	11	0.0	. . .	. . .	31	3.2	1.0	0.1, 7.5
Other*	85	10.6	1.1	0.5, 2.3	33	6.1	2.1	0.5, 9.5
Arrested in county with high chlamydia rate among womenc
Yes	510	10.4	1.3	0.7, 2.2	415	4.1	3.3	0.8, 14.5
No	188	8.5	1.0	Reference	157	1.3	1.0	Reference

Note. CT = Chlamydia trachomatis; CI = confidence interval; DUI = driving under the influence of alcohol.

^aData on age were missing for 7 women.

^bArrest categories not mutually exclusive.

^cRate greater than 1999 California state chlamydia rate among women of 249.9 per 100 000.

* P value for gender comparisons < .0001 (Pearson χ² test).

Chlamydia trachomatis and Neisseria gonorrhoeae Prevalence

Overall, 88 and 3 participants were positive for C trachomatis and N gonorrhoeae, respectively, for an overall C trachomatis prevalence of 6.9% (95% CI = 5.6%, 8.5%) and an overall N gonorrhoeae prevalence of 0.2% (95% CI = 0.1%, 0.7%). The male C trachomatis prevalence was 9.9% (95% CI = 7.8%, 12.3%). The overall female C trachomatis prevalence was 3.3% (95% CI = 2.0%, 5.1%), and 8.9% (95% CI = 2.9%, 22.1%) among women aged 18 to 25 years, 4.3% (95% CI = 1.2%, 10.6%) among women aged 26 to 30 years, and 2.6% among women aged older than 30 years (95% CI = 1.3%, 4.7%). C trachomatis prevalence was highest among African Americans followed by Hispanic inmates among those with known race/ethnicity. For both men and women, prevalence was higher among those from high C trachomatis morbidity counties—10.4% versus 8.5% among males, and 4% versus 1% among females, respectively, but this was not statistically significant.

Only 3 cases of N gonorrhoeae were detected in this population with an overall prevalence of 0.24% (95% CI = 0.05%, 0.69%) in those tested. One man aged 20 years was positive for N gonorrhoeae (0.1%; 95% CI = 0.0%, 0.8%) and was also coinfected with C trachomatis, whereas 2 women aged 25 and 33 years were positive only for N gonorrhoeae (0.4%; 95% CI = 0.0%, 1.3%). Because of the small numbers of N gonorrhoeae infection detected in the sample, further analysis of demographic and behavioral correlates was limited to C trachomatis infections.

Correlates of Chlamydial Infection

Demographic and arrest history correlates of C trachomatis infection are presented in Table 1 ▶. Other than theft, no arrest charge was associated with C trachomatis infection among males. Although men arrested in high C trachomatis morbidity counties were 25% more likely to be infected with C trachomatis, the increase was not statistically significant. For women, younger age was associated with C trachomatis infection. No associations between any arrest charge and C trachomatis infection were seen for women. Among women, the association between being arrested in a county with a high C trachomatis morbidity (rate > 249.9/100 000) and C trachomatis infection was not statistically significant. After adjusting for age among females, the multivariate OR estimates did not differ significantly from the unadjusted estimates (data not shown).

DISCUSSION

This analysis of C trachomatis and N gonorrhoeae prevalence in a population of California prison inmates is 1 of few recent studies on the burden of bacterial STDs for incoming prison populations and the first on inmates entering California prisons. Our findings show that N gonorrhoeae prevalence was low in this population, which was consistent with recent findings from other settings.17^–23 The low N gonorrhoeae prevalence may be attributable to 2 aspects of N gonorrhoeae epidemiology: first, gonococcal infection tends to be more focally distributed in high-prevalence sexual networks, which may not have been captured in this sample, and, second, N gonorrhoeae–infected persons are more likely to be symptomatic and therefore treated before prison entry.

By contrast, C trachomatis prevalence was high, especially among young, asymptomatic men. In California and nationally, there are no C trachomatis screening guidelines for men and the burden of C trachomatis infection in asymptomatic men aged 18 to 26 years has recently been estimated to be between 3.7% and 5.3%,7^,18^–22 albeit the number of published reports are few and settings have varied from military health assessments to population-based surveys. The current findings, which demonstrate comparable prevalence in men and women aged 18 to 25 years, also contrast with C trachomatis and N gonorrhoeae prevalence estimates in jails and juvenile detention facilities where C trachomatis prevalence among women is typically significantly higher than among men; the prevalence among young, asymptomatic male inmates has been reported to be 6% and less.7^,23^–26

Our study had several limitations. Because of cost constraints, only men aged 25 years and younger could be tested and included in this study. Given the high C trachomatis prevalence among this younger population, further examination of the C trachomatis burden among all men entering prison may be warranted.

Furthermore, no behavioral, medical history, or symptom data were available in this unlinked analysis. This prevented any analysis of the proportion of C trachomatis infections that were asymptomatic, as well as the ascertainment of behavioral risk factors that may aid in establishing screening guidelines in prison settings. The prevalence among inmates entering the prison system was measured 5 years ago and may not reflect current prevalence in this population; thus, ongoing prevalence monitoring systems are needed to assess current trends.

Finally, data were unavailable on length of incarceration in the county jail setting before prison entry, as was history of previous C trachomatis screening, diagnosis, and treatment. This prevented us from examining persistent infection or treatment failure for C trachomatis infections acquired or identified before or during incarceration. Therefore, the C trachomatis and N gonorrhoeae infections identified in this study among newly arriving prison inmates may either represent disease that was acquired before arrest or disease acquired while detained in the jail setting.

Despite these limitations, the high prevalence of C trachomatis found at intake in this sample of younger inmates entering California prisons indicates that this is a population at high risk for infection and with minimal access to C trachomatis screening and treatment. Currently, the NCCHC standard states that “as soon as possible, but no later than 7 calendar days, an initial health assessment is completed on each inmate after arrival which includes . . . laboratory and/or diagnostic tests for communicable diseases including sexually transmitted diseases.”12^(p62) Furthermore, the Centers for Disease Control and Prevention 2002 STD Treatment Guidelines recommend the screening of sexually active women aged 25 years and younger in clinical settings27 but make no mention of screening men and screening outside a clinical situation, such as on entry to correctional facilities.

Our findings support age-based screening recommendations that target women aged 30 years and younger. Because men aged older than 25 years were not included in our study, our analysis could not address age-based screening recommendations for men. Further examination of C trachomatis prevalence in men aged 25 years and older may be needed to develop targeted, evidence-based screening recommendations based on prevalence. However, the current NCCHC standard clearly includes women aged 30 years and younger as well as men aged 25 years and younger; despite this standard, screening has not been widely implemented in the correctional setting. One possibility for lack of implementation may be that the NCCHC standard is not specific enough, about targeting younger men and women for STD screening; ages are not specified, which may lower adherence to the standard.

Of note, more than 90% of prison inmates in this study were incarcerated in a county jail before being transferred to a prison. However, few counties routinely screen inmates at entry to jail. Although it is impossible to rule out C trachomatis acquisition in the jail setting, the combination of high prevalence and infrequent jail screening suggests a missed opportunity to decrease STD burden before prison entry. Lastly, the ease of urine-based testing provides an excellent opportunity to identify and treat C trachomatis and N gonorrhoeae infections in any corrections facility, i.e., prisons, jails, or juvenile detention centers, that serve at-risk populations with high prevalence of infection and minimal access to health care services before detention.