In univariate sensitivity analysis, we first varied the proportion of screened men who were from the high partner change group, which changed the positivity in screened men. “Men and women” saved more discounted QALYs than “expand women” as long as the positivity in the screened men exceeded 4.3%. This corresponded to less than 9% of the screened men coming from the high partner change group (compared with 17% at baseline and 5% in the general population). Although screening men with a positivity of 4.3% was as effective in saving discounted QALYs as expanded screening of women with a positivity of 3.4%, it was more costly. To achieve equivalence in discounted total societal cost with “expand women,” the positivity in screened men would have to be 4.7% (Fig. 1).
When varying male test cost, the net societal cost of “women and men” was lower than that of “expand women” unless the male testing cost exceeded $27.50 per male tested. However, male testing cost never exceeded $20,600 per QALY saved over that of “expand women,” which was the incremental cost per QALY when test cost equaled $37.52, the top of the sensitivity analysis range. When the probability of PID in women infected with chlamydia was varied, the net societal cost of “men and women” was lower than that of “expand women” unless the probability of PID in women with chlamydia was less than 8%. However, “women and men” saved more QALYs than “expand women” across the entire range of PID probabilities, costing $210 per QALY saved over that of “expand women” when the probability of PID in women infected with chlamydia was 5%.
For the multiway sensitivity analysis, we varied the variables shown with ranges in Tables 1, 3 simultaneously to compare the cost and QALYs saved of “women and men” with “expand women” (Fig. 2). Positive numbers in the figure indicate that “women and men” was higher in cost or saved more QALYs than “expand Women;” negative numbers indicate the reverse. The average cost difference of the 100 iterations was +$41,300 in cost and +3.93 QALYs saved, indicating that the average incremental cost-effectiveness ratio for “women and men” compared with “expand women” was $10,520 per QALY saved.
The 4-city male chlamydia screening study showed that it is possible to find and screen high-risk men in many venues. Surveillance data and a literature review by Rietmeijer et al. in this issue shows that in a number of settings, principally, detention facilities prevalences are consistently high.3,27 We show that it is possible to screen men in such venues at modest cost per specimen collected. The data collected during PN activities show that PN can be an effective adjunct to screening men, but in the transmission model, most of the reduction in disease in women came from detection and treatment of infection in men, rather than through PN. Most settings lack the resources to undertake PN for partners of patients diagnosed with chlamydia.28 Providing PN for partners of screened men might be more feasible given the relatively small male screening program envisaged by this analysis, but even if PN cannot be provided, our modeling results suggest that screening men for chlamydia can still have a beneficial impact on women.
Screening young women and women with risk factors is an obvious use of chlamydia prevention dollars, because surveillance data and research studies consistently show that these women have the highest prevalence of chlamydia among women.29,30 Less data are published on the prevalence in older women, but what data exist indicate that the prevalence among older women, such as those 30 years of age or older, is much lower than that of younger women.
Screening men for chlamydia is a viable intervention alternative to increasing screening of women if high-risk men (those with high numbers of partners and high chlamydia prevalence compared with the general population of men) can be found; the relative benefit of screening men over screening women from the general population diminishes when the prevalence among screened men declines toward the population average. Similarly, the advantage of screening men diminishes if a group of unscreened women has a chlamydia prevalence that is the same as or even somewhat less than that of the men who are being considered for screening. However, in many instances, men with high chlamydia rates are relatively easily identified–correctional settings are an obvious possibility, and one in which specimens can usually be obtained relatively inexpensively. Another readily accessible population with high chlamydia rates is male military recruits.27,31
The cost advantage of screening high-risk men is impacted by the cost of specimen collection. This suggests that the most advantageous venues for screening high-risk men are those in which men are already receiving a health screening, as they are at jail intake, or in another setting where specimens can be obtained at low cost. Screening men was cost saving compared with expanded screening of women at the baseline because the program cost of the 2 intervention alternatives was equal, and screening men saved more QALYs. The multiway sensitivity analysis suggests that societal-perspective cost savings cannot be assumed to prevail in all instances when screening roughly equivalent numbers of men and women as an add-on to an existing program of screening women. The net societal cost of screening high-risk men averaged about $41,300 over the 5-year period, or $8260 per year, but consistently saved more QALYs than a program slightly expanding screening of women from the general population. The average incremental cost per QALY gained by screening high-risk men was $10,520, which would put it among the most cost-effective preventive interventions recently examined in a systematic review.32
When varying the parameters of the model, the program costs of the 2 interventions will not necessarily be equivalent. The salient finding from the sensitivity analysis is that screening a limited number of high-risk men was found to be consistently more effective in saving QALYs than screening an equivalent number of women from the general population, and that the average cost per QALY of screening the men was in a range typically considered to be cost-effective.
Our findings and the transmission model we used are subject to limitations. First, infection and reinfection were random events controlled for by only 2 factors: population prevalence and rate of partner change. We did not model partnerships explicitly, so treated patients with untreated partners were not more likely to be reinfected than patients whose partners were successfully treated (however, we did control for the likelihood that partners of infected patients were more likely to be infected). Modeling of specific partnerships is not readily accomplished by compartmental models. However, reinfection had no impact on the model that was distinct from an initial infection. Partner treatment did reduce incident infections in the opposite sex generally, but the impact was not modeled by partnership. We did not examine interventions that call for repeat screening of women or men testing positive, although such screening is suggested.6 Second, because individuals in the population were not tracked, we did not account for the possibility that a given partner might be named by more than 1 diagnosed patient; this could have led to an overestimation of the effectiveness of PN. Third, the limitations inherent in any modeling exercise apply to this one, as well: they are simplifications of reality and may not capture important characteristics of transmission. The baseline values for the parameters in Table 1 were selected to be consistent with the literature and to yield an equilibrium prevalence that represents the true prevalence of chlamydia in the population for men and women. The intervention of “women” at baseline yielded postintervention equilibrium chlamydia prevalences in men (2.1%) and women (2.4%) that are close to those found in a nationally representative survey for the same age groups (2.0% in men and 2.5% in women).30
Another population-based model has suggested that screening additional women averts more adverse outcomes of chlamydial infection than screening men.11 However, it assumed population-wide age-based screening of men as an alternative to population-wide age-based screening of women. Our modeling results also show that screening women would likely be more effective than screening men if a broad-based male screening program were undertaken. However, when the prevalence among screened men substantially exceeds that of the additional women who would otherwise be screened, screening the men can be the best approach. In our model, the baseline prevalence of screened men was nearly twice that of women (86% higher). In numerous settings, such a prevalence difference might exist: For example, in California family planning clinics in 2006, the chlamydia positivity was 2.0% for women 30 or older versus 6.4% in women 15 to 24 years (these data are from the Infertility Prevention Program, which targets young, sexually active women under 26 for screening).29 The median chlamydia positivity reported in 2006 among men entering correctional facilities in California was 5.5%.29 If the alternative use of male chlamydia screening funds would go to increase screening among women 30 years of age or older, our findings indicate that screening the men at jail intake would be cost-effective.
In most areas of the US, male screening will likely be an intervention that will only be adopted or implemented to a limited degree. Chlamydia screening rates among young women are low; therefore, screening women will remain the priority.33,34
However, within those limits, targeted screening of high-risk men can play a role in chlamydia prevention.
1. McNabb SJ, Jajosky RA, Hall-Baker PA, et al. Summary of notifiable diseases–United States, 2006. MMWR Morb Mortal Wkly Rep 2008; 55:1–92.
2. Weinstock H, Berman S, Cates W Jr. Sexually transmitted diseases among Am youth: incidence and prevalence estimates, 2000. Perspect Sex Reprod Health 2004; 36:6–10.
3. Centers for Disease Control and Prevention. Sexually transmitted disease surveillance 2006. Atlanta, GA: US Department of Health and Human Services, 2007:166.
4. Paavonen J, Westrom L, Eschenbach D. Pelvic inflammatory disease. In: Holmes KK, Sparling PF, Stamm WE, et al., eds. Sexually Transmitted Diseases. New York, NY: McGraw-Hill, 2008:1017–1050.
5. US Preventive Services Task Force. Screening for chlamydial infection–including ocular prophylaxis in newborns. In: US Preventive Services Task Force, ed. Guide to Clinical Preventive Services. McLean, VA: International Medical Publishing, 2002:325–333.
6. Centers for Disease Control and Prevention, Workowski KA, Berman SM. Sexually transmitted diseases treatment guidelines 2006. MMWR Morb Mortal Wkly Rep 2006; 55:1–94.
7. Elster AB, Kuznets NJ. AMA Guidelines for Adolescent Preventive Services (GAPS). Baltimore, MD: Williams & Wilkins, 1994.
8. Stamm WE. Chlamydia trachomatis
infections of the adult. In: Holmes KK, Sparling PF, Stamm WE, et al., eds. Sexually Transmitted Diseases. New York, NY: McGraw-Hill, 2008:575–594.
9. Gift TL, Owens CJ. The direct medical cost of epididymitis and orchitis: Evidence from a study of insurance claims. Sex Transm Dis 2006; 33:S84–S88.
10. Hogben M, Kissinger P. A review of partner notification for sex partners of US men infected with chlamydia. Sex Transm Dis. In press.
11. Welte R, Kretzschmar M, Leidl R, et al. Cost-effectiveness of screening programs for Chlamydia trachomatis
. Sex Transm Dis 2000; 27:518–529.
12. Roberts TE, Robinson S, Barton PM, et al. Cost effectiveness of home based population screening for Chlamydia trachomatis
in the UK: economic evaluation of chlamydia screening studies (ClaSS) project. BMJ 2007; 335:291.
13. de Vries R, van Bergen JE, de Jong-van den Berg LT, et al. Systematic screening for Chlamydia trachomatis
: Estimating cost-effectiveness using dynamic modeling and Dutch data. Value Health 2006; 9: 1–11.
14. Adams EJ, Turner KM, Edmunds WJ. The cost effectiveness of opportunistic chlamydia screening in England. Sex Transm Infect 2007; 83:267–275.
15. 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. In press.
16. Schillinger JA, Dunne EF, Chapin JB, et al. Prevalence of Chlamydia trachomatis
infection among men screened in 4 US cities. Sex Transm Dis 2005; 32:74–77.
17. Dunne EF, Chapin JB, Rietmeijer CA, et al. Rate and predictors of repeat Chlamydia trachomatis
infection among males. Sex Transm Dis. In press.
19. Gunter DC, Gift TL, Abban B, et al. Bugs, drugs, dollars, and tests: making the most of scarce resources for chlamydia and gonorrhea screening and treatment [Internet]. Paper presented at: 2004 National STD Prevention Conference; Mar 8–11, 2004; Philadelphia, PA. [Cited 2008 Jul 31]. Available at: http://www.cdc.gov/STDConference/2004/Slides/C-sessions/C1/AllPresentors.pps
20. Hethcote HW, Yorke JA. In: Bremermann HJ, Cowan JD, Hirsch W, et al., eds. Gonorrhea Transmission Dynamics and Control. Berlin, Germany: Springer-Verlag, 1984:105.
21. Garnett GP, Mertz KJ, Finelli L, et al. The transmission dynamics of gonorrhoeae: modelling the reported behaviour of infected patients from Newark, New Jersey. Phil Trans R Soc Lond B Biol Sci 1999; 354:787–797.
22. Hu D, Hook EW III, Goldie SJ. Screening for Chlamydia trachomatis
in women 15 to 29 years of age: a cost-effectiveness analysis. Ann Intern Med 2004; 141:501–513.
23. Institute of Medicine. Chlamydia. In: Stratton KR, Durch JS, Lawrence RS, eds. Vaccines for the 21st Century: A Tool for Decisionmaking. Washington, DC: National Academy Press, 1999:149–158.
24. Siegel JE, Weinstein MC, Torrance GW. Reporting cost-effectiveness studies and results. In: Gold MR, Siegel JE, Russell LB, et al., eds. Cost-Effectiveness in Health and Medicine. New York, NY: Oxford University Press, 1996:276–303.
25. Blower SM, Dowlatabadi H. Sensitivity and uncertainty analysis of complex models of disease transmission: an HIV model, as an example. Int Stat Rev 1994; 62:229–243.
26. Consumer price index–all urban consumers [Internet]. Washington, DC: Bureau of Labor Statistics, 2007 [cited 2007 Dec 14]. Available at: http://www.bls.gov/cpihome.htm
27. Rietmeijer CA, Hopkins E, Geisler WM, et al. Chlamydia trachomatis
positivity rates among men in selected venues in the United States. Sex Transm Dis. In press.
28. Golden MR, Hogben M, Handsfield HH, et al. Partner notification: low coverage for gonorrhea, chlamydia, and HIV. Sex Transm Dis 2003; 30:490–496.
29. Centers for Disease Control and Prevention. Chlamydia Prevalence Monitoring Project Annual Report 2006. Atlanta, GA: US Department of Health and Human Services, 2007:116.
30. Datta SD, Sternberg MR, Johnson RE, et al. A population based survey of chlamydia and gonorrhea among persons 14–39 years of age in the United States, National Health and Nutrition Examination Survey (NHANES) 1999–2002. Ann Intern Med 2007; 147:89–96.
31. Nevin RL, Shuping EE, Frick KD, et al. Cost and effectiveness of chlamydia screening among male military recruits: Markov modeling of complications averted through notification of prior female partners. Sex Transm Dis. 2008; 35:705–713.
32. Maciosek MV, Coffield AB, Edwards NM, et al. Priorities among effective clinical preventive services. Am J Prev Med 2006; 31:52–61.
33. National Committee for Quality Assurance. The state of health care quality 2005. Washington, DC, National Committee for Quality Assurance, 2007:74.
34. Levine WC, Dicker LW, Devine O, et al. Indirect estimation of chlamydia screening coverage using public health surveillance data. Am J Epidemiol 2004; 160:91–96.
35. Laumann EO, Gagnon JH, Michael RT, et al. The social organization of sexuality. Chicago, IL: University of Chicago Press, 1994:718.
36. Quinn TC, Gaydos CA, Shpherd M, et al. Epidemiologic and microbiologic correlates of Chlamydia trachomatis
infection in sexual partnerships. JAMA 1996; 276:1737–1742.
37. Oxman GL, Smolkowski K, Noell J. Mathematical modeling of epidemic syphilis transmission. Sex Transm Dis 1996; 23:30–39.
38. Ghani AC, Swinton J, Garnett GP. The role of sexual partnership networks in the epidemiology of gonorrhea. Sex Transm Dis 1997; 24:45–56.
39. Farley TA, Cohen DA, Elkins W. Asymptomatic sexually transmitted diseases: The case for screening. Prev Med 2003; 36:502–509.
40. Kramer MA, Aral SO, Curran JW. Self-reported behavior patterns of patients attending a sexually transmitted disease clinic. Am J Public Health 1980; 70:997–1000.
41. 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.
42. Kretzschmar M, Welte R, Van Den Hoek A, et al. Comparative model-based analysis of screening programs for Chlamydia trachomatis
infections. Am J Epidemiol 2001; 153:90–101.
43. Molano M, Meijer CJ, Weiderpass E, et al. The natural course of Chlamydia trachomatis
infection in asymptomatic Colombian women: A 5-year follow-up study. J Infect Dis 2005; 191:907–916.
44. Morré SA, Van Den Brule AJ, Rozendaal L, et al. The natural course of asymptomatic Chlamydia trachomatis
infections: 45% clearance and no development of clinical PID after one-year follow-up. Int J STD AIDS 2002; 13:12–18.
45. Magid D, Douglas JM Jr, Schwartz JS. Doxycycline compared with azithromycin for treating women with genital Chlamydia trachomatis
infections: an incremental cost-effectiveness analysis. Ann Intern Med 1996; 124:389–399.
46. Washington AE, Johnson RE, Sanders LL Jr. Chlamydia trachomatis
infections in the United States. What are they costing us? JAMA 1987; 257:2070–2072.
47. Wiesenfeld HC, Hillier SL, Krohn MA, et al. Lower genital tract infection and endometritis: insight into subclinical pelvic inflammatory disease. Obstet Gynecol 2002; 100:456–463.
48. Douglas JM Jr, Newman D, Bolan G, et al. Low rate of pelvic inflammatory disease (PID) among women with incident Chlamydia trachomatis
(CT) infection. Int J STD AIDS 2001; 12(suppl 2):65–67 [abstract].
49. Low N, Egger M, Sterne JA, et al. Incidence of severe reproductive tract complications associated with diagnosed genital chlamydial infection: The Uppsala women's cohort study. Sex Transm Infect 2006; 82:212–218.
50. Bachmann L, Richey CM, Waites K, et al. Patterns of Chlamydia trachomatis
testing and follow-up at a University Hospital Medical Center. Sex Transm Dis 1999; 26:496–499.
51. 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:S77–S83.
52. Gaydos CA, Theodore M, Dalesio N, et al. Comparison of three nucleic acid amplification tests for detection of Chlamydia trachomatis
in urine specimens. J Clin Microbiol 2004; 42:3041–3045.
53. Johnson RE, Green TA, Schachter J, et al. Evaluation of nucleic acid amplification tests as reference tests for Chlamydia trachomatis
infections in asymptomatic men. J Clin Microbiol 2000; 38:4382–4386.
54. Van Der Pol B, Ferrero D, Buck-Barrington L, et al. Multicenter evaluation of the BDProbeTec ET system for detection of Chlamydia trachomatis
and Neisseria gonorrhoeae
in urine specimens, female endocervical swabs, and male urethral swabs. J Clin Microbiol 2001; 39:1008–1016.
55. Schachter J, Chow JM, Howard H, et al. Detection of Chlamydia trachomatis
by nucleic acid amplification testing: Our evaluation suggests that CDC-recommended approaches for confirmatory testing are ill-advised. J Clin Microbiol 2006; 44:2512–2517.
56. Howell MR, Quinn TC, Brathwaite W, et al. Screening women for Chlamydia trachomatis
in family planning clinics. Sex Transm Dis. 1998; 25:108–117.
57. Association of Public Health Laboratories. 2001 sexualy transmitted diseases laboratory test method survey [report]. Washington, DC, Association of Public Health Laboratories; 2002. 19p.
58. Begley CE, McGill L, Smith PB. The incremental cost of screening, diagnosis, and treatement of gonorrhea and gonorrhea and chlamydia in a family planning clinic. Sex Transm Dis 1989; 16:63–67.
59. Medical Econonomics Company. 2001 Drug Topics red Book, 105th ed. Montvale, NJ: Medical Economics Company, Inc., 2001 800p.
60. Gift TL, Malotte CK, Ledsky R, et al. A cost-effectivenes analysis of interventions to increase repeat testing in patients treated for gonorrhea of chlamydia at public sexually transmitted disease clinics. Sex Transm Dis 2005; 32:542–549.
61. Yeh JM, Hook EW III, Goldie SJ. A refined estimate of the average lifetime cost of pelvic inflammatory disease. Sex Transm Dis 2003; 30:369–378.
62. Rein D, Kassler WJ, Irwin KL, et al. Direct medical cost of pelvic inflammatory disease and its sequelae: Decreasing, but still substantial. Obstet Gynecol 2000; 95:397–402.
63. Blandford JM, Gift TL. Productivity losses attributable to untreated infection with Chlamydia trachomatis
and to associated pelvic inflammatory disease in reproductive-age women. Sex Transm Dis 2004; 33:S117–S121.
64. Haddix AC, Hillis SD, Kassler WJ. The Cost effectiveness of azithromycin for Chlamydia trachomatis
infections in women. Sex Transm Dis 1995; 22:274–280.
65. Gold MR, Franks P, McCoy KI, et al. Toward consistency in cost-utility analyses. Med Care 1998; 36:778–792.
Each person was classified as susceptible (uninfected), exposed (incubating), infected (symptomatically or asymptomatically), or suffering sequelae (PID or epididymitis). The model is depicted in Figure 1. After infection, persons were assumed to progress through an incubation period to either symptomatic or asymptomatic infection. Chlamydial infections were assumed to either resolve spontaneously (if unscreened and asymptomatic) or through infected persons seeking treatment on their own (if unscreened and symptomatic).
We modeled a representative population, and categorized people based on sex and rate of sex partner change only (we excluded differences in race and age). The population was closed, with no entry or exit, because of the relatively short time horizon. Estimates for the rate of partner change in the population and the per-partnership rate of chlamydia transmission were taken from the literature. We used data provided by men screened in 4 cities for the rates of partner change among the men screened (ci,j, Table 1), proportion of treated men interviewed by DIS (ηi), proportion of partners in the last 60 days (PNI) elicited by DIS (πi,j), proportion of elicited partners located and notified (f), and return rates for treatment (ri,j). For the equivalent female PN program, we used either the same values as for the men or values from the literature (Table 1). For partner notification, we assumed a 10-day lag between index patient screening and partner notification, assuming the partners were locatable at all.
See Table 1 for variable definitions. In the notation below, S = susceptible (uninfected), E = exposed, A = asymptomatically infected, X = symptomatically infected, SQ = sequelae, λ = force of infection, ρ = the sexual mixing coefficient, PN = infected patients treated through PN, SP = infected sex partners, RFA and RFX = recovery from infection if asymptomatic or symptomatic, respectively. In the subscript notation below and in Table 1, i = sex (1,2); j = sexual activity level (1,2 based on low, high); k = sexual activity level of partner (1,2 based on low, high). An accent sign (eg, i') denotes values different from the one under consideration (eg, when i = 1, i' = 2).
The expression for RFX is the same as that for RFA, except that dxi is substituted for dai and PNXij is substituted for PNAij. Sequelae in women (SQ2j) can be either symptomatic or asymptomatic; for modeling purposes the only difference between the 2 is duration (dsq). The sequelae state was modeled as a simplified transition from infection back to susceptibility, in which persons would test negative for chlamydia, but not be immediately susceptible to reinfection. If the sequelae state were instead modeled as one in which persons would test positive, the results in Tables 3a and 3b would change minimally (less than 1%), leaving the ordinal ranking of screening options and dominance unchanged (results available from the authors upon request). The degree of assortative mixing between rate of partner change groups is controlled by equation 13 and the modeling assumption that the number of sex acts for each rate of partner change group balances between men and women.21 When j = k, Δjk = 1; when j ≠ k, Δjk = 0. Given this, if ε = 1 then mixing is fully assortative, meaning low-rate persons always choose other low-rate persons as sex partners; if ε = 0, then mixing is completely random and the partner change groups of sex partners are based on the total partnerships in the population.21 We calculated QALYs by applying published health utility measures to symptomatic chlamydial infections and sequelae, and assumed that sequelae of PID developed at previously estimated rates.9,22,23 These health utility measures, the onset delay (where relevant) and the duration for which they were incurred are shown in Table 1a.
We started the model with 1 infected human in the high partner change group and allowed the system to come to an equilibrium prevalence for men and women, then applied the various intervention alternatives. In the absence of screening, the overall prevalence of chlamydia in women was 2.9% in women and 2.3% in men; under the baseline intervention of screening 35% of women, the equilibrium prevalence of chlamydia was 2.4% in women and 2.1% in men, which is close to that found in a population-based survey.30