Correctional facility inmates are at higher risk for sexually transmitted infections with rates considerably higher than those of the general population.1 In women, chlamydia (CT) and gonorrhea (GC) can cause serious and costly sequelae, including pelvic inflammatory disease, which can lead to chronic pelvic pain, ectopic pregnancy, and infertility. Estimates in the literature have indicated that screening and treating women for these infections so as to avert sequelae are cost-effective compared with symptom-based testing.2 The Centers for Disease Control and Prevention (CDC) and US Preventive Services Task Force provide guidelines and recommendations for screening women for these infections.3 The CDC does not provide similar guidelines for screening of male adults because the sequelae of male infection (epididymitis) are not severe.4 The US Preventive Services Task Force gives male screening for CT an “I” (incomplete evidence) rating, and cost-effectiveness analyses of male screening have reached varying conclusions.5 However, studies have estimated probable benefits of screening men in high-prevalence settings such as in jails.6 Although 2 studies associated jail-based male screening programs with a reduction in disease burden in the community7,8 another study of a Philadelphia program found no such evidence.9
A pilot screening study conducted by the Arizona Arrestee Reporting Information Network (AARIN) indicates high prevalence rates of CT and GC infections among men and women in Maricopa County (Phoenix area) jails.10 Among male arrestees, 7% and 4.6% were tested positive for CT and GC, respectively, and among female arrestees, 10% and 5% were tested positive for CT and GC, respectively (see Table 1 for age-based rates). The reported cases of CT and GC in the general population in year 2010 were approximately 0.41% and 0.06% of Maricopa residents, respectively.10,11
The Maricopa County Correctional system provides medical care to more than 130,000 persons per year, with annual operating costs of approximately $50 million per year.12 The Maricopa County Correctional Health Services (CHS) currently screens women arrestees 35 years and younger for CT and GC infections during the time of physical examination (PE). Male arrestees are not screened because of the unavailability of evidence of screening benefits and limited resources. However, the high prevalence of infections among male arrestees in Maricopa jails evidenced in the AARIN study indicates that a large population of women could be exposed to untreated male infections and thus be put at risk for serious sequelae. We used available data on the Maricopa jail population, including local prevalence data from the AARIN study and demographics data from the Maricopa county jails, to simulate infection in jail inmates and transmission to female partners upon their release during 1 calendar year. We estimated the number of infections averted in women by screening of male inmates in Maricopa County jails compared with the current strategy of symptom-based testing and treating. We also estimated the costs incurred for such a screening program and cost-effectiveness as the cost per infection averted in women.
We developed a stochastic individual simulation model (in NetLogo 4.1.3 software) by individually simulating each male arrestee from the time of entry into jail to either, release from jail if person is not infected at the time of release (both uninfected and infected but treated before release), or until recovery from infection if person is infected at the time of release. We simulated 100,000 male inmates, which was approximately the number of inmates entering jails in Maricopa County each year. The time-unit in the simulation is a calendar day. Age at entry into jail (Table 1) and length of stay in jail (distributed as 39%, 50%, 52%, 62%, 74%, and 99% stay in jail <1, 2, 3, 8, 14, and 365 days, respectively) for each person in the simulation were assigned to match those of male inmates in the Maricopa County jails. At the time of entry, the proportion of arrestees infected with CT, GC, or both was simulated to match the age-based prevalence of these infections reported in the AARIN pilot study. We assumed that there were no new infections during the duration of stay in jail. For those who were released from jail with infection, we simulated possible transmissions to women as follows. Upon release from jail, assuming a heterosexual population, each person was assigned a female partner and variables that defined the sexual relationship, such as duration of partnership, number of sex acts, and condom use. When a relationship ended, a person could have a new partner, and the sexual behavioral variables were reassigned with different values to define the new partnership. During each time-unit after release from jail, the probability that the infected inmate could transmit the infection to his uninfected female partner was calculated using the Bernoulli equation.
where p is the probability of transmission per time-unit, α and β are the probabilities of transmission per unprotected and protected sex act, respectively, n the number of sex acts for that time-unit, and f the proportion of protected sex acts. Transmission beyond that of infected persons released from jail to uninfected female partners was not modeled. See Tables 1 and 2 for a list of data parameters and assumptions used in the model. To model stochasticity among individual persons, where applicable, parameter values were assigned by drawing random numbers from probability distributions. The number of sex acts per person per year was uniformly distributed. For each new sexual partnership of a person, duration of the partnership was exponentially distributed with mean estimated using number of partners per year. The number of sex acts per partnership was assigned proportional to the duration. Condom use was distributed by age. The probability of transmission per sex act was determined based on a uniform distribution. The duration of infection was also uniformly distributed.
We simulated a population of 100,000 male inmates each under 5 scenarios: (1) symptom-based testing: this was the baseline and is equivalent to the current scenario where male inmates who seek medical help based on symptoms are tested for CT and GC; (2) screen all during PE: this scenario proposes screening all male arrestees for CT and GC during the routine PE offered to all inmates, which usually occurs between days 8 to 14 from the time of entry to jail; (3) screen all inmates 35 years or younger during PE: given that the prevalence of the infections was lower in the older population of jail inmates (Table 1) and to reflect current CDC screening recommendations for women in correctional facilities, this scenario proposes screening only male arrestees who are 35 years or younger during the routine PE; (4) screen all on days 2 to 3: given that approximately 62% to 74% of inmates are released from jail by the time of PE, this scenario proposes screening all male arrestees for CT and GC on the second or third day from the time of entry into jail; and (5) screen all inmates 35 years or younger on days 2 to 3: this scenario proposes screening only male arrestees who are 35 years or younger on the second or third day from the time of entry to jail.
In all scenarios, we made the following assumptions. Testing would be done using a urine-based combination assay for both CT and GC. We assumed that 73% of all inmates in jail at the time of screening would agree to test, which is the proportion in Maricopa County jails accepting a PE. It takes approximately 2 to 3 days to receive test results and approximately 5 to 7 days from time of test before an inmate who is tested positive receives treatment. All inmates who are tested positive and are in jail were provided with and accepted the necessary treatment. Infected inmates who were released from jail before this time were followed up and, if found (with a probability of 80% [expert opinion]), were treated. Chlamydia would be treated with 1 dose of azithromycin and GC or coinfection with 1 dose of azithromycin plus 1 dose of ceftriaxone. We kept track of test and treatment costs incurred by each inmate, which are covered by the CHS or by the Maricopa County Department of Public Health (MCDPH) (Table 3). In each scenario, we also assumed that infected inmates who were not tested while in jail but developed symptoms after release from jail would seek medical help, that is, get tested and treated outside the jail. We assumed that those costs would be incurred by a different entity outside of MCDPH or CHS and hence were not included in our analyses. Testing costs, treatment costs, and personnel wages for follow-up were provided by Maricopa County Health Department in 2011 dollars, and a microcosting direct measurement technique was used to estimate cost inputs. To derive the labor costs for follow-up and treatment, we multiplied the staff time associated with each activity by the compensation, that is, wages plus benefits.
We test for the sensitivity of results on age-based proportion accepting screening, age-based length of stay in jail, earlier availability of treatment for those infected, and additional cost of early screening, that is, on days 2 to 3 instead of with PE (Table 4).
Under each scenario, using the simulation model, we estimated the number of infections in women, that is, transmissions from infected male inmates after release from jail, and the testing and treatment costs incurred by the CHS and MCDPH. For every scenario, we ran the simulation 30 times each with 100,000 inmates to obtain mean and confidence interval values of the results. All costs are in 2011 dollars. We estimated the average cost-effectiveness of scenarios 2 through 5 by dividing the net cost of a scenario by the net number of infections averted compared with scenario 1. By arranging scenarios in order of effectiveness, that is, decreasing order in number of new infections in women, we also estimated incremental cost-effectiveness ratios (ICERs) as incremental cost per infection averted in pairwise comparisons of a strategy with the next most effective strategy and eliminated weakly dominated scenarios. A scenario is weakly dominated if its ICER is greater than that of a more effective scenario, and hence, if all scenarios are feasible and acceptable, weakly dominated scenarios can be eliminated. We then reestimated ICERs among the remaining scenarios. In “Results,” we round estimated values to the nearest 10th unit.
In all scenarios, of the 100,000 inmates entering jail each year, approximately 10,330 were estimated to be infected with CT or GC at the time of entry (Table 5). Under the baseline scenario (scenario 1), where male inmates were tested and treated for CT and GC only if they sought medical help because of occurrence of symptoms, approximately 2150 infected inmates had received treatment by time of release from jail, whereas the rest of the 8160 inmates were still infected and unaware of their infection at the time of release. Upon release from jail, these infected inmates transmitted the infection to approximately 6090 women. Scenario 1 cost approximately $114,960, of which, 50% was incurred by CHS and the remaining 50% by MCDPH.
Under the scenario where all male inmates were offered CT and GC screening along with their PE (scenario 2), which occurred anywhere between days 8 and 14 from the time of entry to jail, approximately 2600 infected inmates had received treatment by time of release from jail, whereas the rest of the 7740 remained infected and unaware of their infection at the time of release. The infected inmates comprised those who were released before testing day or did not accept the test (6930) and those who were tested but released from jail before the test results became available (810). These male inmates transmitted the infection to approximately 5530 women. When only inmates 35 years or younger were offered screening along with their PE (scenario 3), of the infected male inmates, approximately 2570 were tested and treated before release, approximately 7050 infected male inmates left jail before screening day or declined testing, and the remaining 740 infected male inmates were tested but had left the jail before receiving results. These infected men transmitted the infection to approximately 5600 women. Therefore, although a smaller population was screened in scenario 3 compared with scenario 2, given the lower prevalence of CT and GC and data indicating lower sexual risk behavior among older male population than younger men, there was not much difference in the number of transmissions. Scenarios 2 and 3 cost approximately $802,540 and $535,930, respectively. In scenario 2, testing and treatment incurred 92% and 8% of total cost, respectively. In scenario 3, testing and treatment incurred 88% and 12% of total cost, respectively. In both scenarios, CHS and MCDPH incurred 59% and 41% of the total cost, respectively.
When all inmates were offered the test on days 2 to 3 from he time of entry to jail instead of delaying until PE day (scenario 4), there were approximately 7140 inmates with infection at the time of their release from jail. The released infected inmates comprised approximately 5730 inmates who had either left the jail before test day or refused to get tested and had 1,410 inmates who were tested but released before their test results were available. When only inmates 35 years or younger were offered the test on days 2 to 3 from the time of entry to jail (scenario 5), there were approximately 7250 with infection at the time of release, including approximately 1280 who had been tested but had not yet received test results by the time of release. Infected male inmates in scenarios 4 and 5 transmitted the infection to approximately 5000 and 5090 women, respectively. Scenarios 4 and 5 cost approximately $1,249,460 and $819,740, respectively. In scenario 4, testing and treatment incurred 92% and 8% of total cost, respectively. In scenario 5, testing and treatment incurred 89% and 11% of total cost, respectively. In both scenarios, CHS and MCDPH incurred 59% and 41% of the total cost, respectively.
Compared with baseline, scenarios 2 and 3 prevented approximately 560 and 490 cases of infection in women, respectively, whereas scenarios 4 and 5 prevented approximately 1100 and 995 cases, respectively. Average cost-effectiveness, that is, the average cost per infection averted compared with the baseline (scenario 1), was approximately $1240, $860, $1030, and $710 in scenario 2, 3, 4, and 5, respectively (Table 5). Pairwise comparisons to estimate ICERs indicated that scenarios 2 and 3 were weakly dominated and, hence, were eliminated (Table 6). Pairwise comparisons among the remaining scenarios resulted in ICERs of approximately $710 and $4130 per infection averted in scenarios 5 and 4, respectively, compared with their next effective alternative (Table 6).
Sensitivity analyses indicate that scenario 5 has the least cost per infection averted compared with baseline in all cases, except for when the additional cost of screening early (on days 2–3) is $7 per inmate, in which case scenario 3 has the least cost (Fig. 1). Infections averted in women were the highest when there was a faster turnaround of test results and treatment availability (Table 7).
Estimates indicate that offering CT and GC screening for male inmates in Maricopa jails followed with treatment for those infected could avert a large number of infections in women that could have been transmitted by infected male inmates released from jail. Although screening all male inmates averted the most cases of infections compared with screening only younger inmates, it was also relatively costly. Results suggest that screening only inmates 35 years or younger has the potential to avert a considerable number of infections in women. Only approximately 26% to 38% of inmates stay in jail longer than 8 to 14 days (time of PE) and, as such, implementing the screening as soon as possible following arrest should be considered. Screening early, within 2 to 3 days when 48% to 50% are still in jail, averted twice the number of infections. Screening male inmates 35 years or younger on days 2 to 3 of entry to jail has the least cost per infection averted compared with symptom-based testing if early screening has no additional costs or is less than $7 per inmate screened. If early screening costs an additional $7 or more per inmate screened then screening inmates 35 years or younger on PE day had the least cost per infection averted compared with symptom-based testing.
The model is subject to certain limitations. We estimated costs of testing and treatment of CT and GC only and did not include costs averted from sequelae in men or infection and sequelae in women. We estimated only first-level transmission and did not include any of the downstream transmissions from the infected women to others. We also did not include the potential impact of partner services. We assumed sexual behavior equivalent to the general population, but it is likely that some of the inmates are involved in higher-risk behavior. All of the above could have underestimated costs averted and infections averted. Also, individual programs may face different costs for specimen collection or testing and lead to results that differ from what we found.
Public health programs and correctional health systems are frequently faced with priority versus resource challenges. Our estimates indicate that screening male inmates for CT and GC could avert a considerable number of infections and hence possibly prevent cases of costly sequelae in women. These results are consistent with studies on screening at similar high-prevalence settings.13,14 Averting these infections could possibly translate to reducing disease burden in the community, as indicated in some studies.7,8 Another study that did not find any such evidence showed an unaccounted decrease in infections in the community in both control and case groups.9 The costs incurred per infection averted are within range of costs incurred in other screening programs that indicate that screening of men for CT and GC is cost-effective in high-prevalence settings.5,13 Our results indicate that, from the perspective of MCDPH and CHS, screening male inmates 35 years or younger has a lower incremental cost per infection averted compared with screening male inmates of all ages, which is consistent with CDC’s guidelines for screening women.4 Early screening and treatment availability could avert the most infections. In conclusion, in addition to screening women for early detection of CT and GC for prevention of severe sequelae, screening men in high-prevalence settings such as jails could prevent the occurrence of these infections in a large number of women.
1. Hammett TM. Sexually transmitted diseases and incarceration. Curr Opin Infect Dis 2009; 22: 77–81.
2. Roberts TE, Robinson S, Barton P, et al.Screening for Chlamydia trachomatis
: A systematic review of the economic evaluations and modelling. Sex Transm Infect 2006; 82: 193–200.
3. Screening for chlamydial infection: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2007; 147: 128–134.
4. Workowski KA, Berman S. Sexually transmitted diseases treatment guidelines, 2010. MMWR Recomm Rep 2010; 59: 1–110.
5. Gift TL, Blake DR, Gaydos CA, et al.The cost-effectiveness of screening men for Chlamydia trachomatis
: A review of the literature. Sex Transm Dis 2008; 35 (suppl 11): S51–S60.
6. Owusu-Edusei K Jr, Gift TL, Chesson HW, et al.Investigating the potential public health benefit of jail-based screening and treatment programs for chlamydia. Am J Epidemiol 2013; 177: 463–473.
7. Broad J, Cox T, Rodriguez S, et al.The impact of discontinuation of male STD screening services at a large urban county jail: Chicago, 2002–2004. Sex Transm Dis 2009; 36 (suppl 2): S49–S52.
8. Barry PM, Kent CK, Scott KC, et al.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 (suppl 2): S22–S28.
9. Peterman TA, Newman DR, Goldberg M, et al.Screening male prisoners for Chlamydia trachomatis
: Impact on test positivity among women from their neighborhoods who were tested in family planning clinics. Sex Transm Dis 2009; 36: 425–429.
13. Kraut-Becher JR, Gift TL, Haddix AC, et al.Cost-effectiveness of universal screening for chlamydia and gonorrhea in US jails. J Urban Health 2004; 81: 453–471.
14. Gift TL, Lincoln T, Tuthill R, et al.A cost-effectiveness evaluation of a jail-based chlamydia screening program for men and its impact on their partners in the community. Sex Transm Dis 2006; 33 (suppl 10): S103–S110.
15. Chandra A, Mosher WD, Copen C, et al.Sexual behavior, sexual attraction, and sexual identity in the United States: Data from the 2006–2008 National Survey of Family Growth. Natl Health Stat Report 2011; 1–36.
16. Herbenick D, Reece M, Schick V, et al.Sexual behaviors, relationships, and perceived health status among adult women in the United States: Results from a national probability sample. J Sex Med 2010; 7 (suppl 5): 277–290.
17. Reece M, Herbenick D, Schick V, et al.Condom use rates in a national probability sample of males and females ages 14 to 94 in the United States. J Sex Med 2010; 7: 266–276.
18. Stamm WE. Chlamydia trachomatis
infections in the adult. In: Holmes KK, Sparling PF, Stamm WE, et al. eds. Sexually Transmitted Diseases. New York, NY: McGraw-Hill; 2008: 575–593.
19. Garnett GP, Mertz KJ, Finelli L, et al.The transmission dynamics of gonorrhoea: Modelling the reported behaviour of infected patients from Newark, New Jersey. Philos Trans R Soc Lond B Biol Sci 1999; 354: 787–797.
20. Kretzschmar M, van Duynhoven YT, Severijnen AJ. Modeling prevention strategies for gonorrhea and chlamydia using stochastic network simulations. Am J Epidemiol 1996; 144: 306–317.
21. Farley TA, Cohen DA, Elkins W. Asymptomatic sexually transmitted diseases: The case for screening. Prev Med 2003; 36: 502–509.
22. Quinn TC, Gaydos C, Shepherd M, et al.Epidemiologic and microbiologic correlates of Chlamydia trachomatis
infection in sexual partnerships. JAMA 1996; 276: 1737–1742.
23. Warner L, Newman DR, Austin HD, et al.Condom effectiveness for reducing transmission of gonorrhea and chlamydia: The importance of assessing partner infection status. Am J Epidemiol 2004; 159: 242–251.
24. Handsfield HH, Lipman TO, Harnisch JP, et al.Asymptomatic gonorrhea in men. Diagnosis, natural course, prevalence and significance. N Engl J Med 1974; 290: 117–123.
25. Hethcote HW, Yorke JA. Gonorrhea Transmission Dynamics and Control: Lecture Notes in Biomathmatics. New York: Springer; 1984.
26. Schachter J, Hook EW, Martin DH, et al.Confirming positive results of nucleic acid amplification tests (NAATs) for Chlamydia trachomatis
: All NAATs are not created equal. J Clin Microbiol 2005; 43: 1372–1373.
27. Dize L, Agreda P, Quinn N, et al.Comparison of self-obtained penile-meatal swabs to urine for the detection of C. trachomatis
, N. gonorrhoeae
and T. vaginalis
. Sex Transm Infect 2012; 89: 305–307.
28. Koumans EH, Black CM, Markowitz LE, et al.Comparison of methods for detection of Chlamydia trachomatis
and Neisseria gonorrhoeae
using commercially available nucleic acid amplification tests and a liquid pap smear medium. J Clin Microbiol 2003; 41: 1507–1511.
29. Geisler WM. Management of uncomplicated Chlamydia trachomatis
infections in adolescents and adults: Evidence reviewed for the 2006 Centers for Disease Control and Prevention sexually transmitted diseases treatment guidelines. Clin Infect Dis 2007; 44 (suppl 3): S77–S83.
© Copyright 2013 American Sexually Transmitted Diseases Association
30. Barbee LA, Kerani RP, Dombrowski JC, et al.A retrospective comparative study of 2-drug oral and intramuscular cephalosporin treatment regimens for pharyngeal gonorrhea. Clin Infect Dis 2013.