Chlamydia and Gonorrhea Screening Using Urine-Based Nucleic Acid Amplification Testing Among Males Entering New York City Jails

Introduction

Chlamydia trachomatis (Ct) and Neisseria gonorrhoeae (Ng) sexually transmitted infections (STIs) are highly prevalent among jail populations (Division of STD Prevention, 2000; Glaser & Greifinger, 1993; Joesoef, Kahn, & Weinstock, 2006; Kahn et al., 2005; Mertz, Voigt, Hutchins, Levine, & The Jail STD Prevalence Monitoring Project, 2002). Females are disproportionately affected by these STIs and, if untreated, may face serious sequelae including pelvic inflammatory disease, adverse pregnancy outcomes, and infertility (Centers for Disease Control and Prevention [CDC], 1991). Male complications of untreated STI, though less common, include urethritis, epididymitis, and infertility (Stamm, 1999). A 2002 report from the National Commission on Correctional Health Care recommends routine screening of all newly incarcerated detainees for Ct and Ng STIs (National Commission on Correctional Health Care, 2002). While some jails routinely screen females for these STIs, few routinely screen males and often rely on detainee request or symptoms report for STI testing (Parece, Herrera, Voigt, Middlekauff, & Irwin, 1999). However, most males with Ct and/or Ng STIs are asymptomatic (Handsfield, Lipman, Harnisch, Tronca, & Holmes, 1974; Stamm, 1999) and thus go undetected in jails.

Jail-based male STI screening can be an important adjunct to local and national STI prevention and control efforts. Incarcerated males frequently report high-risk behaviors such as unprotected sex or sex with multiple partners in the community and engage in these behaviors soon after jail discharge (Glaser & Greifinger, 1993; Grinstead et al., 2005; Morrow & Project START Study Group, 2009). Jail populations are also less likely to seek routine medical care and STI screening in the community (Conklin, Lincoln, & Tuthill, 2000). Without timely diagnosis and treatment, infected males can unknowingly transmit STI to their sex partners. Jail settings, therefore, provide a public health opportunity to disrupt disease transmission by implementing routine male STI screening and treatment. The availability of noninvasive, highly sensitive urine-based nucleic acid amplification testing (NAAT) has made such STI screening feasible (Akduman, Ehret, Messina, Ragsdale, & Judson, 2002; Chan, Brandt, Olienus, Antonishyn, & Horsman, 2000; Johnson et al., 2002; Van Der Pol et al., 2001).

New York City (NYC) jails make up the second-largest jail system in the United States. The NYC Department of Correction (DOC) operates the facilities, and the Correctional Health Services (CHS) bureau of the NYC Department of Health and Mental Hygiene (DOHMH) is responsible for health care delivery. In 2003, only newly incarcerated females were offered STI screening at admission. Thus, the STI prevalence among newly incarcerated males was unknown. In 2003, CHS conducted a Ct and Ng STI screening pilot study for all newly incarcerated adolescent and adult males using a urine-based NAAT. Its aims were to determine the STI prevalence in a complete male intake cohort, identify predictors associated with STI positivity, determine the treatment rate achievable during incarceration, and assess the feasibility of routine admission, or intake, STI screening.

Method

Results

During the study period, 2,559 males were newly incarcerated and of these, 2,417 (94.5%) met study inclusion criteria. Fourteen males were incarcerated more than once, and 128 excluded males had incomplete medical intake and/or LET/NAAT results due to early jail discharge. No refusals of urine submission occurred. Discussions with clinical staff revealed that medical intake was minimally affected by incorporation of screening activities associated with the study. Urine specimen processing, administration of the supplemental questionnaire and related administrative tasks were incorporated into staff’s existing schedules without deployment of additional personnel or extending process time frames. DOC reports confirmed that at no time did medical intake exceed its allotted 4 hours as a result of study activities.

Median age at intake of the study population was 32 years (interquartile range [IQR]: 23 to 40), with 34% aged ≥ 25 years (Table 1). Non-Hispanic Blacks and Hispanics accounted for 87% of the population. Fifty percent of males resided in the Bronx and Brooklyn; 4% were homeless. Median LOS was 13 days (IQR: 4 to 63).

Of 2,300 (95.2%) detainees with available supplemental questionnaire data on recent STI risk and history, 47 (2.0%) reported an STI and 191 (8.3%) had ≥ 1 new sex partners. Of 2,281 males for whom the data were available, 462 (20.3%) reported 0 total sex partners, 1,366 (59.9%) reported 1, 405 (17.8%) reported 2 to 5, 30 (1.3%) reported 6 to 10, and 18 (0.8%) reported ≥ 11. Less than 1% reported a sex partner with Ct, Ng, syphilis, or HIV. Of 727 (31.6%) detainees reporting unprotected sex, 710 (97.7%) also reported ≥ 1 sex partners and 103 (14.2%), having a new sex partner.

Of 2,417 males, 162 (6.7%) tested positive for Ct/Ng STI (range: 1.5% [age ≥ 41 years] to 12.8% [age 21–25 years]; see Table 1). Positivity for Ct and Ng was 6.5% (95% CI [5.4, 7.5]) and 0.9% (95% CI [0.5, 1.3]), respectively; 0.6% were coinfected. Ng prevalence ranged from 0.5% (age ≥ 36) to 1.5% (age 21 to 25). Ninety percent of cases were aged ≤ 35 (60.0% of total study population) and had a Ct/Ng STI positivity of 10.0%. Ct/Ng-positive males were younger than the overall study population (median: 23 years; IQR: 20 to 28; p < .001).

Seventeen (10.5%) of 162 Ct/Ng-positive males tested LET positive. Of 159 Ct/Ng-positive males with supplemental questionnaire data, 4 (2.5%) reported recent urethral symptoms; 138 (86.8%) were neither LET positive nor symptomatic. The sensitivity, specificity, and positive predictive value of urethral symptoms for Ct/Ng positivity were 2.5% (95% CI [0.8, 6.7]), 98.4% (95% CI [97.7, 98.8]), and 10.3% (95% CI [3.3, 25.1]), respectively, and for LET, 10.5% (95% CI [6.4, 16.5]), 97.5% (95% CI [96.7, 98.1]), and 23.0% (95% CI [14.3, 34.5]), respectively.

STI treatment was administered to 102 (63.0%) of all Ct/Ng-positive males prior to jail release. Treatment courses included azithromycin (n = 91), ceftriaxone (n = 22), doxycycline (n = 10), and ciprofloxacin (n = 9); those dually infected or receiving empiric therapy received treatment for both Ct and Ng STI. Treated males had a greater median LOS than those untreated (median 31 days, IQR: 7 to 80, vs. median 2.5 days, IQR: 1 to 4, respectively; p < .001). Two thirds of untreated infected males were released ≤ 3 days after intake, with 95% discharged ≤ 8 days. Court appearances, legal counsel meetings, other medical appointments, and security, housing, and transport-related issues also hindered timely treatment. One Ct/Ng-positive male refused treatment. In addition, 37 (1.6%) of Ct/Ng-negative males were presumptively treated; of these, 1 (2.7%) was symptomatic, 25 (67.6%) were LET positive, 3 (8.1%) were both, and 8 (21.6%) had other abnormalities on urine dipstick testing (e.g., hematuria). Treatment courses consisted of azithromycin (n = 27), ceftriaxone (n = 22), doxycycline (n = 9), and ciprofloxacin (n = 11); antibiotic information was incompletely recorded for 5 males receiving empiric therapy.

In univariate analysis, Ct/Ng positivity was associated with age (p < .001), LET (p < .001), unprotected sex (p = .01), number of recent sex partners (p < .001), and race/ethnicity (p = .03; see Table 2). Facility (not shown) was associated with Ct/Ng positivity due solely to the adolescent facility (positivity = 11.8%); as such, facility was not included in multivariate analysis.

Adjusted analysis showed age to be a strong predictor of Ct/Ng positivity (Table 2), with the odds of being Ct/Ng positive being greater for all age groups ≤ 35 years compared with the referent group of ≥ 41 years (range: < 21, OR = 9.1; 95% CI [4.3, 19.2] to 31 to 35, OR = 2.9; 95% CI [1.2, 7.0]). When continuous age was examined, the odds of being Ct/Ng positive decreased 9.0% by each increasing year of age. Ct/Ng positivity was also independently associated with positive LET (OR = 5.4; 95% CI [2.9, 10.0]). Overall, no association was found with increasing number of sex partners (OR = 0.6, 95% CI [0.6, 1.0]), but males reporting ≥ 11 recent sex partners had a significantly higher odds of having STI as compared with males reporting no recent partners (OR = 4.2, 95% CI [1.2, 15.0]). When modeling was repeated with Ct-only positivity as the dependent variable, the relationships of predictor variables remained consistent (data not shown).

Direct unit costs of NAAT and complete treatment were as follows: NAAT, $18.40; azithromycin, $18.14; ceftriaxone, $1.32; ciprofloxacin, $0.13; and doxycycline, $0.56. The total direct costs of NAAT and treatment for the study population were calculated as (a) $18.40 × 2,417 NAAT tests = $44,473 + (b) $1,671 for treatment of 102 Ct/Ng-positive males + (c) $512 for presumptive treatment of 37 Ct/Ng-negative males on initial suspicion of STI = $46,656. The overall direct cost per case detected was $46,656/162 = $288 and for cost per case detected and treated, $46,656/102 = $457.

Discussion

To our knowledge, this study was the first to assess the prevalence of Ct and Ng STI in a complete cohort of newly incarcerated males in a U.S. adult jail setting. Our overall Ct/Ng prevalence of 6.7% is consistent with the published reports of substantial STI prevalence among incarcerated male populations (Joesoef et al., 2009; Kahn et al., 2005; Mertz et al., 2002; Schillinger et al., 2005; Trick et al., 2006) and was mostly attributed to Ct infection. Age was the strongest predictor of Ct/Ng STI. Males aged 25 or younger had the highest STI prevalence, as found in other studies (Joesoef et al., 2009; Satterwhite, Joesoef, Datta, & Weinstock, 2008; Schillinger et al., 2005); males aged 26 to 30 and 31 to 35 also had a substantial STI prevalence. Although, generally, Ct/Ng STIs are most prevalent among adolescent and young adult males, our study supports findings by others that incarcerated males older than 25 years also have high STI burdens and could be important to target in jail-based screening (Barry and colleagues, 2007; Nguyen et al., 2004).

In jails without NAAT availability and with short inmate LOS, positive LET and/or urethral symptoms may inform a clinician’s decision to empirically treat for suspected STI. However, we found LET sensitivity and positive predictive value for Ct/Ng STI to be low. While some studies have reported that LET may be useful for STI detection in adolescent males or as a pre-NAAT screening tool for symptomatic males, our findings, like those of others, suggest that the utility of LET as a stand-alone Ct/Ng STI screening method is limited (Blake, Lemay, Gaydos, & Quinn, 2005; Coble, Nordahl-Akesson, Vinnerberg, & Kihlstrom, 2006; Wood, Gaydos, McKee, & Gaydos, 2007). We found that 87% of males with NAAT-identified Ct/Ng STI had no signs or symptoms of infection. Detainees either may not recognize their STI symptoms or may be reluctant to disclose such information to jail clinicians. LET- and symptom-based STI screening strategies for incarcerated males in jails are thus likely to miss most cases of Ct/Ng STI.

With the exception of a very high number of recent sex partners, none of the supplemental questionnaire responses were associated with STI positivity. Recall bias or fear/unwillingness to self-report risk behaviors in jail may have affected the responses. Thus, self-report of risk behaviors may not be a useful parameter for targeted STI screening at jail admission but may continue to inform risk reduction education, counseling, and disease prevention efforts in jails. We found that newly incarcerated males are at high risk for STI. Nearly one third of the study population reported recent unprotected sex; and 20% reported two or more sex partners. While less than 1% of males reported having a recent sex partner with STI or HIV, this finding may reflect unwillingness to disclose the STI or HIV status to a provider or a lack of awareness of the status of their partners. Sex partners may engage in high-risk behavior, including unprotected sex, without mutual awareness of STI and HIV status or partner concurrency (Dawson et al., 1994; Drumright, Gorbach, & Holmes, 2004; Witte, El-Bassel, Gilbert, Wu, & Chang, 2010). This unawareness may lead to a falsely low perception of STI risk and resulting failure to seek routine STI screening in the community.

National recommendations continue to prioritize the STI screening and treatment of females (Division of STD Prevention, 2007; U.S. Preventive Services Task Force, 2005, 2007). However, there is increasing recognition of the value of age-based STI screening of incarcerated males (Barry et al., 2007; Gift, Gaydos, et al., 2008). The CDC recommends that, while screening females remains the priority, programs already screening or planning to screen males for Ct should target newly incarcerated males aged under 30 years (Division of STD Prevention, 2007). Barry et al. (2007) similarly suggest that, where funding is limited, STI control programs should prioritize screening of incarcerated females in juvenile detention and those aged 30 and younger, followed by newly incarcerated males aged 25 and younger.

In our study, 63% of infected males were treated prior to jail release; the majority of untreated males were discharged before NAAT results became available. Follow-up for infected males with short LOS remains a challenge; detainees in our study may not have had valid, complete, or current residential addresses on record. Public health and correctional authorities, including discharge planners, community correctional officers, health providers, and local STI control programs, should collaborate to identify ways to facilitate treatment of STI-positive persons and their sex partners after release from jail.

Incorporation of STI screening into the extant medical intake proved to be feasible and acceptable. Because urine was routinely collected for dipstick testing at intake, the addition of NAAT did not require additional specimen collection. The supplemental questionnaire was easily incorporated into the routine taking of medical histories. Importantly, the study protocol did not adversely affect overall intake processing times, allowing detainees to be housed within the required DOC time frames. While superior to LET- and symptom-based screening, urine-based NAAT carries a high direct monetary cost, which may limit its use in other jail settings (Broad et al., 2009). Our direct cost per STI case detected was $288 and cost per case treated was $457. A cost-effectiveness evaluation was beyond the scope of this study; however, others have found that screening incarcerated males for Ct/Ng STI may be cost effective (Gift, Blake, Gaydos, & Marrazzo, 2008; Kraut-Becher, Gift, Haddix, Irwin, & Greifinger, 2004; Rietmeijer, Hopkins, Geisler, Orr, & Kent, 2008). Correctional facilities considering male STI screening should work with regional STI control programs to formulate effective screening policies based on local STI epidemiology, competing health-based priorities, and available resources (Barry et al., 2007).

In 2005, CHS instituted routine urine-based Ct/Ng NAAT screening at intake to NYC jails for males aged 35 and younger, while maintaining opportunistic screening for older males. Unpublished data from the year following implementation showed that 10% of more than 39,000 male incarcerations tested were Ct/Ng infected, similar to our study’s finding; 70% of infected males were successfully treated. That year, NYC jail Ct and Ng case reporting surpassed by 40% that of the 10 DOHMH STI clinics combined, and, as a result, citywide-reported case rates of male Ct and Ng infection increased by 59% and 4%, respectively (Pathela et al., 2009).

This study had limitations. First, as noted above, questionnaire responses may not have accurately characterized the true STI risk history in this population and symptoms may have been underreported due to unawareness of STI status or reluctance to disclose such sensitive information. Studies suggest that the mode of question administration and situational context may have an impact on the quality of self-reported risk behavior (Brener, Billy, & Grady, 2003; Weinhardt, Forsyth, Carey, Jaworski, & Durant, 1998). Other studies have suggested overreporting of sexual behaviors and drug and alcohol use by adolescents (Alexander, Somerfield, Ensminger, Johnson, & Kim, 1993; Winters, Stinchfield, Henly, & Schwartz, 1991). We were unable to ascertain the extent to which under- or overreporting of risk behaviors occurred in our study. Second, we did not question detainees about gender or sexual orientation of sex partners and recognize that STI prevalence may have varied by these factors. Third, study exclusion due to incomplete medical intake may have introduced selection bias, although the percent excluded (5%) was small relative to the final population. Fourth, our study was conducted in December, near the holidays, when new jail admissions are typically lower than in other months; thus our observed Ct and Ng prevalence may be underestimates. Finally, we did not evaluate data on previous incarcerations. NYC jails have highly recidivist populations, and the number and/or duration of incarcerations may have had an impact on study results due to differential access to health care and time spent incarcerated.

In summary, this pilot study showed that a urine-based NAAT STI screening program for newly incarcerated adolescent and adult males in NYC jails was feasible and detected a substantial Ct/Ng STI prevalence that otherwise would have been undiagnosed. Age was the strongest predictor of positivity. The majority of STI-positive males were treated prior to jail release, thus potentially reducing transmission to female sex partners. Our findings suggest that jails are important venues for age-based targeted male Ct/Ng STI screening.