PEOPLE WHO HAVE BEEN DIAGNOSED with chlamydia or gonorrhea represent a high-risk group for reinfection. A recent study of patient-delivered partner therapy for women diagnosed with chlamydia in family planning, adolescent, sexually transmitted disease (STD), and primary care clinics or emergency or other hospital departments found that 4-month reinfection rates ranged from 12% to 15%.1 The Centers for Disease Control and Prevention (CDC) recommend advising women with chlamydia to be rescreened 3 to 4 months after treatment.2 The U.S. Preventive Services Task Force has not made a formal recommendation for rescreening infected patients, but notes that rescreening women may be valuable because of high rates of reinfection.3
The Gonorrhea Community Action Project (GCAP), a feasibility study funded by the CDC and the National Institute of Mental Health (NIMH), assessed several methods to encourage STD clinic patients diagnosed with chlamydia or gonorrhea to return for a repeat screening 3 months after treatment. This clinic-based intervention was a part of a larger study that analyzed gonorrhea prevention interventions targeted at STD clinic patients, healthcare providers, and members of community-based organizations in three communities (Central Harlem in New York City, NY, southwest Los Angeles, California, and Prince George’s County, Maryland).4
Materials and Methods
The GCAP clinic intervention has been described in detail previously.4 Briefly, eligible patients were men and women between 14 and 30 years of age who were diagnosed with and treated for chlamydia or gonorrhea at either the Prince George’s County STD Clinic (Prince George’s County, Maryland) or the Ruth Temple Health Center STD Clinic (Los Angeles County, California). Both clinics serve high-prevalence communities.4 Enrollment began in April 2000 and concluded in July 2001. Eligible clinic clients provided written informed consent and were randomized to one of three intervention arms. In intervention 1, patients were given a verbal recommendation to return for a chlamydia and gonorrhea screening in 3 months and given a reminder card with five to seven reasons for returning. This was the standard of care (standard treatment). In intervention 2, patients were given a brief verbal recommendation and reminder card, but also informed they would receive a $20 incentive on their return (incentive). In intervention 3, patients received a verbal recommendation to return and reminder card, but also participated in a one-on-one 15- to 25-minute interview and counseling session with a trained interviewer that helped patients assess their own risk, identify barriers to returning, and provided the patient with reasons and motivations for returning. Patients in intervention 3 were also given a reminder phone call and mailed a letter 3 months after the counseling session to encourage their return (motivational counseling and reminder). In each intervention, patients who returned to the clinic provided a urine specimen that was screened for chlamydia and gonorrhea using a ligase chain reaction (LCR) nucleic acid amplification assay (LCx; Abbott Laboratories, Abbott Park, IL). Symptomatic patients and patients with positive test results were treated according to CDC guidelines.2 Table 2 lists the number of participants for each intervention arm from each site. There were no significant differences between the sites for any demographic or sexual behavior variables.4
The return rate among study participants in intervention 3 was higher than among those in intervention 1 or 2. For each intervention, there were no significant differences in return rates between the two sites or between men and women. To assess whether the significantly higher return rate for intervention 3 was attributable to the counseling or to the reminder, a second study began in Los Angeles in July 2001, which also consisted of three arms.4 Intervention 4 was the same as intervention 1 in all respects (standard treatment). Intervention 5 consisted of a brief verbal recommendation and reminder card plus a phone call reminder (reminder). Intervention 6 had the components of intervention 3 except that participants were not given either a letter or phone call to encourage their return for repeat screening (motivational counseling). Both phases of the study were approved by Institutional Review Boards at the CDC and at the individual study sites (Maryland Department of Health and Mental Hygiene, CA State University–Long Beach, and County of Los Angeles Department of Health Services). Enrollment ended in June 2002.
Cost Measurement and Estimation
The cost variables used in the analysis are in Table 1A. Study staff who were specifically trained in time-motion methods and who used data collection forms that were uniform across the two sites tracked the amount of time necessary for clinic workers to deliver each of the interventions, including time for preparation, patient counseling activities, supervision, and quality assurance. Included in supervision and quality assurance time were direct supervision activities, chart review, review of audiotapes of the sessions, and employee meetings relevant to the interventions. During the study, study supervisory staff reviewed the audiotapes from the motivational interview sessions conducted as part of interventions 3 and 6. Initially, 100% of counseling sessions were monitored, but this was later reduced to 10%. Staff also recorded the amount of time required to place phone calls and prepare reminder letters. Staff collected wage and fringe benefit rate data for workers involved in the intervention; these were combined with the time expenditures to determine the labor costs associated with each intervention. Counseling staff and supervisors attended a 2-day workshop for training in the first three clinical interventions in September 1999. The cost of this training was not incorporated into the per-patient costs reported in the results.
We also estimated the costs of sequelae from infection with chlamydia or gonorrhea. The total cost of sequelae included the direct medical costs of treatment plus the lost productivity costs attributable to the sequelae. For women, the primary sequelae considered were pelvic inflammatory disease (PID) and its sequelae of chronic pelvic pain, infertility, and ectopic pregnancy. Literature estimates of the likelihood of sequelae and their lifetime costs were used. Averted costs were discounted at 3% per year.10,13–15 Productivity costs were also taken from the literature.16,17 For men, the sequelae considered were epididymitis and orchitis; we used literature-based estimates of their likelihood and cost for both the direct medical costs and productivity losses.10–13
We estimated the cost for supplies, postage, and office and counseling space used in the intervention (including supervisory and quality assurance activities) based on average commercial rental rates per square foot in the areas served by the clinics for commercial office space.18 Costs for utilities were approximated based on the average cost per square foot for commercial buildings in the appropriate region of the country.19
Patient time costs were estimated using mean income data for persons 15 to 29 years old.20 These wage data were adjusted by race, ethnicity, and sex-specific labor force participation rates to match the participants in the study.4 In addition to the time spent in the initial counseling, we assessed 2 hours of lost productivity (adjusted by the labor force participation rate) plus $5 in travel-related costs for each clinic visit.20 We applied this cost to both repeat screening and treatment visits.
All other costs such as those for sequelae, testing and treatment, and the cost of clinic visits for repeat screening and treatment if the repeat screening test was positive were taken from the literature.5–8,10–15
Cost-Effectiveness Analysis Perspectives
We conducted the analysis from two perspectives: program and societal. In the program-perspective analysis, only the intervention costs were considered. These included those for staff and overhead for the initial counseling, as well as direct medical costs related to testing and treatment of acute infection in patients who returned to the clinic. All sequelae and patient lost productivity and travel costs were omitted. In the societal perspective analysis, all costs were included, including those reflected in the program-perspective analysis, those attributable to sequelae of chlamydial and gonococcal infection, and those incurred by patients for healthcare visits and lost productivity. The only exception is that the incentive payments for returning patients in intervention 2 were counted in the program-perspective analysis but not in the societal-perspective analysis, because the net cost to society for the payments was zero. The program lost the money, but the patients gained it. Also, the incentives represented a transfer payment and were not in exchange for goods or services.
To determine the cost per infection treated for each perspective, we divided the costs described here by the number of infections detected in returning patients using patient return data from the study. For ease of comparison, we then calculated all of these values per 1,000 initial patients counseled.
Data on clinic breakdown by sex, return rates, and rates of infection among those returning were directly measured in the study (Table 1B).4 Test performance parameters were taken from the literature.31–34 We also used a literature-based estimate of the probability that a returning patient would have symptoms or signs if reinfected with chlamydia or gonorrhea, and we assumed that patients who were symptomatic would be treated presumptively and not recalled for treatment after test results. We assumed asymptomatic patients would be recalled by the clinic for treatment if their test results were positive; we used literature estimates of the rate of return for treatment.24–26
We used interventions 1 and 4 (the interventions closest to the existing standard of care) as the baseline to determine the number of infections averted for each intervention. We estimated the number of infections among the returnees using literature estimates of the sensitivity and specificity of LCR on urine specimens.31–34 We applied these values to the positivity rate for chlamydia and gonorrhea among those who returned for screening to estimate the prevalence of each infection. To estimate the number of infections in patients who did not return for screening, we assumed the prevalence of each infection was the same as in those who returned for screening and that patients were not tested or treated for the subsequent infections.
Sensitivity Analysis and Additional Analysis Considerations
To facilitate sensitivity analysis, a decision tree was constructed using DATA Professional version 1.0 (TreeAge Software, Williamstown, MA). Sensitivity analysis ranges are provided in Table 1 (A and B). We determined the sensitivity analysis ranges by calculating the exact binomial 95% confidence intervals when we had sufficient data to do so.
For sensitivity analysis related to supervisory time associated with quality assurance, we determined the upper bound by calculating the 95% confidence intervals as described here based on observed supervisory and quality assurance time (which included monitoring 10% of counseling sessions). The lower bound was calculated by assuming that 5% of sessions would be monitored.
The return rates for the interventions are reported separately by patient gender in Table 1B. The sensitivity analysis ranges for the return rates are the 95% confidence intervals for the average return rates in each intervention for both genders combined adjusted for patient symptom status as described in the next paragraph. Because interventions 1 and 4 were identical in design, they were combined for the purposes of this analysis. Although the return rate for intervention 4 was lower than for intervention 1, the difference was not significant at P <0.05 At baseline, the gender-specific rates were used, but because the rates did not differ by gender at P <0.05 (using Fisher exact test) for any of the interventions, we examined the impact of using a combined average rate in sensitivity analysis. Because many of the interventions did not have significantly different return rates, we also examined the impact of setting the return rates equal for all of the interventions across the range of return rates in Table 1B. However, the return rate for intervention 5 (reminder) was significantly different from the return rates for interventions 1 and 4 (standard treatment; Fisher exact P <0.01) and intervention 2 (incentive; Fisher exact P <0.05). The return rate for intervention 3 (motivational counseling and reminder) was significantly different from that of interventions 1 and 4 (Fisher exact P <0.05).
To facilitate comparisons between the interventions, we calculated incremental cost-effectiveness ratios, which we determined by ranking the interventions from least to most effective using the number of chlamydial or gonococcal infections treated through repeat screenings as the outcome. We chose this outcome because the intervention combined men and women, making outcomes such as cases of PID averted or cases of epididymitis or orchitis averted impractical (however, as noted previously, PID, epididymitis, and orchitis costs were included in the societal-perspective calculations). The incremental cost-effectiveness ratios were calculated by determining the difference in total intervention cost between two interventions when applied to 1,000 index patients divided by the difference in chlamydial and gonococcal infections treated between the two interventions. When calculating the ratios, we omitted interventions that were dominated by the next most effective intervention (i.e., the omitted interventions were both less expensive and more expensive in terms of the cost per infection treated than the next most effective intervention).
The timeframe covered by our cost-effectiveness analysis would be the amount of time necessary to counsel 1,000 patients, which would depend on patient volume. We assumed all costs were linear (in other words, that the cost per patient counseled would be the same regardless of patient volume, all else being equal). The analytic horizon of the cost-effectiveness analysis was 10 years, by which time we assumed all sequelae costs would be incurred. All costs were adjusted for inflation to reflect 2001 prices using the Medical Care component of the Consumer Price Index for All Urban Consumers (CPI-U).35
Table 3 shows the weighted average return rates for each intervention (for both genders combined) as well as the number of chlamydial and gonococcal infections treated and intervention costs, standardized to a hypothetical cohort of 1,000 patients (given the gender breakdown in Table 1B, the patient mix would be 442 women and 558 men). Table 3 also shows the incremental cost per additional infection detected through repeat screenings performed for patients who returned. The results in Table 3 are based on baseline values in Tables 1A and 1B. The interventions maintain the same rank order in terms of overall cost and cost per infection treated regardless of perspective. The cost per patient counseled was lowest for interventions 1 and 4, regardless of perspective, although for intervention 2, it was very close to that of interventions 1 and 4 (only approximately 3% higher when considering total societal costs). When considering the cost per infection treated, intervention 5 was least costly, regardless of perspective. Intervention 5 was also the most effective, having the highest return rate for men and women.
Interventions 3 (motivational counseling and reminder) and 6 (motivational counseling) were dominated (i.e., they were more expensive in terms of total cost and also less effective) by interventions 2 (incentive) and 5 (reminder).
When chlamydia and gonorrhea prevalences were varied over the ranges in Table 1B, intervention 5 (reminder) remained the least costly in terms of cost per infection treated. At the lowest chlamydia prevalence examined, 3%, the total societal cost per infection treated for intervention 5 was $4,092 compared with $6,136 for intervention 3 (motivational counseling and reminder), the next most-effective intervention. Intervention 5 was also the most effective and the least costly in terms of cost per infection treated when the return rates for the other interventions were set at the top end of the ranges shown in Table 1B. These results were true for both the program and societal perspectives.
When the return rates for all five interventions were set equal to each other and varied across a range of 10% to 50% for both genders combined, they broke into two groups in terms of total societal cost: Interventions 1 and 4 (standard treatment), 2 (incentive), and 5 (reminder) were all within approximately 1% of each other at the low end of cost, with intervention 5 the most expensive of the group. Interventions 3 (motivational counseling and reminder) and 6 (motivational counseling) were 9% to 13% more costly than the low-cost interventions.
The effects of the sensitivity analysis on supervision and quality assurance time can be seen in Table 4. This table shows the cost per infection treated for both perspectives at both the baseline supervision and quality assurance costs and at the lower-end supervision and quality assurance costs from Table 1A. Although the cost per patient counseled, cost per returnee, and cost per infection treated decreased for all interventions relative to the costs in Tables 2A and B, the magnitude of the changes varied. In the program-perspective analysis, cost per infection treated dropped by 19% (for intervention 5) to 31% (for interventions 1 and 4). In the societal-perspective analysis, the reduction in intervention cost per infection treated was from 15% (for intervention 5) to 26% (for interventions 1 and 4). At the lower-end costs for supervision and quality assurance, the incremental cost-effectiveness ratios remained approximately the same, and the interventions that were dominated by others (interventions 3 and 6) remained dominated.
Other sensitivity analyses such as return rates for the interventions when varied individually affected the overall cost and effectiveness of each of the interventions, but did not create important threshold values.
In terms of cost per patient counseled, all of the interventions were more expensive than the baseline (interventions 1 and 4, standard treatment). The per-patient costs (for patients initially counseled, patients returning, and infections treated) can easily be calculated from Table 3 by dividing by the appropriate denominator. Intervention 5 (reminder) was markedly more expensive in terms of per patient counseled (regardless of perspective) than interventions 1 and 4, but the cost per infection treated was approximately half that of interventions 1 and 4. Much of the cost of intervention 5 was incurred because of its effectiveness—returning patients led to costs for testing and treatment that were not incurred by the less-effective interventions. Although more effective interventions averted more sequelae costs, this effect was muted in our analysis because we considered both men and women together. However, even when considering women alone, none of the interventions would be cost-saving. The cost per infection treated will vary depending on the prevalence of chlamydia and gonorrhea in returning patients; at the prevalence rates that were estimated among returnees, the program cost per infection treated (approximately $500 for intervention 5 at baseline prevalences) was comparable to that of some screening programs in other venues.36–38
We did not incorporate the effect of treatment on transmission to future sex partners. It is also possible that the patients who returned may have been more disposed to seek regular STD screening or other health care as a result of the interventions. Both of these considerations could be additional benefits (as well as program costs) of these interventions that were not modeled.
Because none of the interventions was cost-saving, looking at the relative societal costs of the interventions when the return rates are equivalent provides a useful additional way to compare them. If the interventions were equally effective at encouraging patient return, interventions 1 and 4 (standard treatment) would still be least costly. However, intervention 5 (reminder) would only be approximately 1% more expensive overall. This is because, as Table 1A shows, the counseling cost (either from the program or societal perspective) for the two interventions was similar. Because the reminder intervention had a significantly higher return rate than the standard treatment or incentive interventions, this would suggest that the reminder intervention might be a better choice than the other two, assuming that the additional case detection and treatment that it produces is worth the additional cost. To the extent it is more costly, it is so because it accomplishes what it is designed to do better than the alternatives.
Another consideration is that the return rate among men for intervention 2 (incentive) was high, but it was not effective in encouraging women to return (Table 1B). Because repeat screening recommendations typically focus on women, this would suggest that other alternatives may be worthwhile to consider.
The level of quality assurance and other administrative and supervisory activities in this research project were higher than would be experienced in actual practice. For example, the CDC guidelines for HIV counseling (which is similar in complexity to the motivational counseling used in this project, if not more so) recommend review of taped sessions twice monthly for the first 6 months of a counselor’s employment, then monthly for the next 6 months, and quarterly thereafter.39 This would be a less intense and less costly level of quality assurance than was used in GCAP. Therefore, the low supervision and quality assurance cost columns of Table 4 may reflect costs that might be typical of the basic level of quality assurance recommended by the CDC for a similar intervention.
We did not include the one-time training costs in the costs reported in Tables 3 and 4. These costs would be apportioned over the number of sessions that counselors would complete before retraining became necessary or before they stopped counseling (as a result of assuming other duties or employee turnover). During the 26-month GCAP project period, retraining was not needed. Had training costs been apportioned among the total number of patients enrolled in the intervention, the additional labor cost per patient counseled would have been approximately $3.35. Although this would increase the intervention cost, total cost would still fall if training costs were considered along with lower supervision and quality assurance: program-perspective cost per infection treated would drop by 17% to 25%, and societal-perspective cost per infection treated would drop between 8% and 11%, depending on the intervention (results not shown).
The limitations and caveats discussed previously4 apply also to the findings of this cost-effectiveness analysis. The cost per patient counseled is subject to variation depending on the amount of supervision and quality assurance that would be needed to ensure adequate intervention delivery and the rate of sequelae in treated and untreated repeat infections. Another limitation is that some patients may have become reinfected, developed symptoms, and sought care at a point other than the study clinic. This would make the actual “return rates” higher than those observed in the study and would change the cost per infection treated, particularly when considering the societal perspective.
It is striking that the cost per infection treated (from either perspective) was lowest for the most effective intervention. Although intervention 3, the motivational counseling and reminder intervention also had return rates significantly greater than interventions 1 and 4 (standard treatment); its higher cost per patient counseled makes intervention 5, the reminder intervention, a more cost-effective option.
1. Schillinger JA, Kissinger P, Calvet H, et al. Patient-delivered partner treatment with azithromycin to prevent repeated Chlamydia trachomatis
infection among women. Sex Transm Dis 2003; 30:49–56.
2. Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines 2002. MMWR 2002; 51(RR):1–80.
3. US Preventive Services Task Force. Screening for chlamydial infection: recommendations and rationale. Am J Prev Med 2001; 20(suppl):90–94.
4. Malotte CK, Ledsky R, Hogben M, et al. Comparison of methods to increase repeat testing in persons treated for gonorrhea and/or chlamydia at public sexually transmitted disease (STD) clinics. Sex Transm Dis 2004; 31:637–642.
5. Howell MR, Quinn TC, Brathwaite W, Gaydos CA. Screening women for Chlamydia trachomatis
in family planning clinics. Sex Transm Dis 1998; 25:108–117.
6. Association of Public Health Laboratories. 2001 Sexually Transmitted Diseases Laboratory Test Method Survey [Report 1–19]. Washington, DC: Association of Public Health Laboratories, 2002:1–19.
7. Begley CE, McGill L, Smith PB. The incremental cost of screening, diagnosis, and treatment of gonorrhea and chlamydia in a family planning clinic. Sex Transm Dis 1989; 16:63–67.
8. 1998 Drug Topics Red Book. Montvale, NJ: Medical Economics Co, Inc, 1998.
9. Handsfield HH, Stamm WE. Treating chlamydial infection: compliance versus cost. Sex Transm Dis 1998; 25:12–13.
10. 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.
11. Washington AE, Johnson RE, Sanders LL Jr. Chlamydia trachomatis
infections in the United States: What are they costing us? JAMA 1987; 257:2070–2072.
12. Gift TL, Owens C. The direct medical cost of epididymitis and orchitis: evidence from a study of insurance claims [Abstract]. ISSTDR Conference; July 27–30, 2003; Ottawa, Canada. Sex Transm Dis (in press).
13. Institute of Medicine. Chlamydia. In: Stratton KR, Durch JS, Lawrence RS, eds. Vaccines for the 21st Century: A Tool for Decisionmaking. Washington, DC: Academy Press, 1999:149–158.
14. Rein D, Kassler WJ, Irwin KL, Rabiee L. Direct medical cost of pelvic inflammatory disease and its sequelae: Decreasing, but still substantial. Obstet Gynecol 2000; 95:397–402.
15. Shafer M-AB, Pantell RH, Schachter J. Is the routine pelvic examination needed with the advent of urine-based screening for sexually transmitted diseases? Arch Pediatr Adolesc Med 1999; 153:119–125.
16. Blandford JM, Gift TL. The productivity losses attributable to untreated chlamydial infection and associated pelvic inflammatory disease in women. Sex Transm Dis (in press).
17. Washington AE, Katz P. Cost of and payment source for pelvic inflammatory disease trends and projections. JAMA 1991; 266:2565–2569.
18. Oncor International. North American Office Market Report [Report]. Washington, DC: Oncor International, 2002:1–4.
19. US Bureau of the Census. Statistical Abstract of the United States, 121st ed. Washington, DC: US Government Printing Office, 2001.
20. Bureau of Labor Statistics. Current population survey: March 2002 supplement [database on the Internet]. Washington, DC: Bureau of Labor Statistics [updated April 5, 2002, cited July 8, 2004]. Available at: http://www.bls.gov/cps
21. 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.
22. Farley TA Cohen DA, Elkins W. Asymptomatic sexually transmitted diseases: The case for screening. Prev Med 2003; 36:502–509.
23. Oh MK, Smith KR, O’Cain M, Kilmer D, Johnson J, Hook EWI. Urine-based screening of adolescents in detention to guide treatment for gonococcal and chlamydial infections. Arch Ped Adolesc Med 1998; 152:52–56.
24. Hook EW III, Spitters C, Reichart CA, Neumann TM, Quinn TC. Use of cell culture and a rapid diagnostic assay for Chlamydia trachomatis
screening. JAMA 1994; 272:867–870.
25. Foglia GM, Rhodes P, Goldberg M, St. Louis ME. Completeness of and duration of time before treatment after screening women for Chlamydia trachomatis
infections. Sex Transm Dis 1999; 26:421–425.
26. Schwebke JR, Sadler R, Sutton JM, Hook EW III. Positive screening tests for gonorrhea and chlamydial infection fail to lead consistently to treatment of patients attending a sexually transmitted disease clinic. Sex Transm Dis 1997; 24:181–184.
27. Haddix AC, Hillis SD, Kassler WJ. The Cost effectiveness of azithromycin for Chlamydia trachomatis
infections in women. Sex Transm Dis 1995; 22:274–280.
28. Douglas JM Jr, Newman D, Bolan G, et al. Low rate of pelvic inflammatory disease (PID) among women with incident Chlamydia trachomatis
(CT) infection [Abstract]. Int J STD AIDS 2001; 12(suppl 2):65–67.
29. Rees E. The treatment of pelvic inflammatory disease. Am J Obstet Gynecol 1980; 138:1042–1047.
30. Centers for Disease Control and Prevention. Recommendations for the prevention and management of Chlamydia trachomatis
infections. MMWR 1993; 42(RR):1–39.
31. Black CM. Current methods of laboratory diagnosis of Chlamydia trachomatis
infections. Clin Microbiol Rev 1997; 10:160–184.
32. Black CM, Marrazzo JM, Jones RB, et al. Head-to-head comparison of cervical and urine LCR and PCR with a DNA probe test for Chlamydia trachomatis
(CT) infection in women [Abstract]. STIs at the Millenium–Past, Present, and Future; May 3–5, 2000; Baltimore, MD.
33. Koumans EH, Johnson RE, Knapp JS, St. Louis ME. Laboratory testing for Neisseria gonorrhoeae
by recently introduced nonculture tests: A performance review with clinical and public health considerations. Clin Infect Dis 1999; 27:1171–1180.
34. Stary A, Ching S, Teodorowicz L, Lee H. Comparison of ligase chain reaction and culture for detection of Neisseria gonorrhoeae
in genital and extragenital specimens. J Clin Microbiol 1997; 35:239–242.
35. Bureau of Labor Statistics. Consumer price index—all urban consumers [Report, cited December 1, 2003]. Available at: http://www.bls.gov/cpihome.htm
36. Gift TL, Varghese B, Kraut J, et al. The cost-effectiveness of jail-based STD and HIV prevention programs and their impact on inmate and community health [Abstract]. National STD Prevention Conference; March 4–7, 2002; San Diego.
37. Kraut-Becher JR, Gift TL, Haddix AC, Irwin KL, Greifinger RB. Cost-effectiveness of universal screening for chlamydia and gonorrhea in US jails. J Urban Health 2004; 81:453–471.
38. Mehta SD, Bishai D, Howell MR, Rothman RE, Quinn TC, Zenilman JM. Cost-effectiveness of five strategies for gonorrhea and chlamydia control among female and male emergency department patients. Sex Transm Dis 2002; 29:83–91.
39. Rogers MF, Fowler MG, Lindegren ML. Revised guidelines for HIV counseling, testing, and referral. MMWR 2001; 50(RR):1–62.