All costs are reported as year 2000 US dollars and are considered from the government perspective (see Table 1). Costs are adjusted using the consumer price index when necessary. In the year 2000, the Malawian government purchased a 2-g dose from its Central Medical Stores in Lilongwe for $0.05 (Malawi MoH). Men with persistent symptoms who return for additional treatment represent an additional cost to the government. Estimates for the costs to the government for a single outpatient clinic visit to a Malawian government hospital range from less than $1 up to $2 to $3.24 STD management tends to be rapid. We adopt a conservative estimate of $0.50 for staff time and supplies and allow this to vary for sensitivity analysis. For men returning because of untreated T. vaginalis, we assume they are not going to receive a second regimen of gentamicin and doxycycline but only metronidazole.25 No overhead or recurring expenses beyond the cost of metronidazole are incurred in this program; however, we do consider a cost of $200 (or $0.02 per man screened) for reprinting the reference materials to reflect the change in management schemes (Malawi MoH).
We do not consider the future costs of HIV infections attributable to T. vaginalis. Any additional expenditure averted on the government's part for HIV-related care would only serve to enhance the cost-effectiveness of the proposed treatment. Thus, this exclusion is conservative.
A standard decision tree was constructed to consider these data using DATA 4.0 (Treeage Software, Inc., Williamstown, MA). The model is applied to a population of 10,000 semiurban men seeking STD treatment, roughly equivalent to the number of men seeking STD treatment annually in Lilongwe. Currently, the only men who receive metronidazole are the asymptomatic male partners of treated women (Table 3, status quo). We estimated the cost and the probability of HIV transmission and/or acquisition attributable to infection with T. vaginalis in each of the following treatment programs of increasing coverage: (A) including men with urethritis for treatment with metronidazole to the status quo, (B) including men with urethritis and with GUD, and (C) including all male attendees of the STD clinic (see Table 3). These values were entered into the Microsoft Excel 2000 software program. The incremental cost-effectiveness ratio (ICER) was calculated by dividing the incremental costs incurred by each respective program by the cases of HIV averted relative to the status quo.
Univariate sensitivity analysis was conducted using the high or low value for each variable shown in Table 1 while holding the other variables at their base values. To vary the transmission and/or acquisition values, each component used to calculate the attributable risk for T. vaginalis shown in Table 2 was allowed to vary while holding the other components at their base value. We selected 4 transmission values to use in the univariate analysis, the very high, high, low, and very low scenarios shown in Table 1. Multivariate sensitivity analysis was conducted to consider the “worst-case” (eg, high cost, low transmission, low T. vaginalis prevalence, high partner notification and/or treatment, low reinfection, high treatment-seeking behavior) and “best-case” (eg, low cost, high transmission, high T. vaginalis prevalence, low partner notification and/or treatment, high reinfection, low treatment-seeking behavior) scenarios.
HIV Transmission Attributable to Infection With Trichomonas vaginalis
Currently, Malawian syndromic STD guidelines for men provide metronidazole initially to only approximately 12% of male clinic attendees, who are the partners of women managed for trichomoniasis (see Table 1). The prevalence of male trichomoniasis in this STD clinic is approximately 20%.25 We estimate the base transmission rate of HIV from the male index case to a susceptible partner to be 0.0370 and the acquisition rate of an HIV-negative index case to be 0.0154 over the duration of infection with T. vaginalis. In the absence of other STDs (ie, they are properly managed), we anticipate the transmission and acquisition rates to increase to 0.0684 and 0.0315 in the presence of untreated trichomoniasis. Subtracting the initial rate from the second rate, the attributable risk for the transmission or acquisition of HIV attributable to trichomoniasis is 0.0314 or 0.0161, respectively (see Table 1).
Assuming a 10% reinfection rate and that one third of untreated men seek appropriate treatment in a timely manner, we estimate that persistent male trichomoniasis is responsible for 27.2 new cases of HIV, including susceptible male index cases (acquisition) and susceptible female partners (transmission) for every 10,000 men who receive government care for their STD (see Table 3). As we expand coverage for male trichomoniasis, we see a reduction in the number of new HIV infections attributable to untreated T. vaginalis infection, from 18.6 to 7.3 to 4.5.
We estimate that current efforts to treat T. vaginalis under the national program cost $361.80 per 10,000 men attending STD clinics (see Table 3). Program costs increase across the interventions under consideration, as treatment is given to more men, from $639.10 per 10,000 men when metronidazole is added for men presenting with urethritis (scheme A) to $711.80 for complete coverage (scheme C). Hence, incremental costs over the current status quo range from $277.50 to $350.20.
Cost-Effectiveness of Expanding Sexually Transmitted Disease Services to Include Treatment for Male Trichomoniasis
Complete coverage (scheme C) represents an excess cost over the status quo of $350.20 and prevents 22.7 cases of HIV compared with the status quo, translating to an ICER of $15.43 (see Table 3). The small number of cases of HIV transmission observed even with 100% coverage can be attributed to reinfections. Covering most clinic attendees (plan B: asymptomatic referrals plus urethritis and GUD, representing nearly 90% of male clinic attendees) averts 19.9 cases of HIV and provides an ICER of $15.85. Because there is a high prevalence of infection with T. vaginalis in men with GUD in this clinic, just adding men with urethritis averted fewer incident cases of HIV (8.6 cases), with an ICER of $32.27.
Under univariate sensitivity analysis, the HIV transmission events averted ranged from 2 to 52. Model results were robust to partner notification and treatment, secondary treatment-seeking behavior, and metronidazole efficacy but were more sensitive to other parameters.
Results were most sensitive to variation in the probability of HIV transmission. The very high transmission scenario, accounting for extreme variations in number of sexual partners reported by the enrolled men, provides the most favorable (lowest) ICERs. Under treatment plans A, B, and C, we see the highest number of cases averted, 19.8, 45.6, and 52.0, respectively, over the current treatment protocol, yielding more favorable ICERs (Table 4). Setting the duration of infection to 1 month produced our very low transmission scenario. Here, we observe the lowest number of cases of HIV averted as compared to the status quo, from 2.2 to 5.0 to 5.7 per treatment plan. Attendant ICERs were the largest we observed during the univariate sensitivity analysis.
Our results were moderately sensitive to the prevalence of T. vaginalis, reinfection rates, and the cost of a return visit. Using a low value for the prevalence of T. vaginalis observed in an asymptomatic male population from a neighboring clinic8,12 increases ICERs (see Table 4), whereas a 35% prevalence of trichomoniasis26 significantly reduces ICERs because of an increased number of cases averted (18.1, 36.2, and 42.9 cases across plans A, B, and C, respectively) associated with a modest decrease in costs relative to our base analysis. Increasing reinfection rates lowered the effectiveness of the treatment schemes, because fewer cases were averted (4.7, 11.0, and 12.2 across plans A, B, and C, respectively) with costs remaining steady. Revisit costs represent a large component of the overall costs incurred. Therefore, decreasing and increasing the revisit costs resulted in changes to the ICERs attributable entirely to fluctuations in cost.
We created a worst-case scenario and a best-case scenario by minimizing and optimizing all variables used in the sensitivity analysis. For transmission and acquisition values, we used the high and low scenarios rather than the very high and very low values. High attributable transmission and acquisition attributable to T. vaginalis, high T. vaginalis prevalence, high partner notification and treatment rates, high treatment-seeking behavior rates, low revisit costs, and low reinfection rates averted 14.3, 28.5, and 33.8 cases of HIV across plans A, B, and C, respectively, at an overall cost savings (see Table 4). The converse situation provided ICERs that were 5 to 7 times those seen in the base case analysis because of the modest numbers of HIV infections averted (2.4, 4.8, and 5.7 infections across plans A, B, and C, respectively) and significant cost increases.
HIV now infects more than 40 million individuals, with greater than 95% of new cases occurring in the developing world, most of which are attributable to heterosexual sex.1 In areas where the prevalences of STDs are also high, the cofactor effect may account for a considerable burden of excess transmission. Even in areas with a high-prevalence mature-stage HIV epidemic, STDs may still account for a substantial number of new HIV cases, all the while contributing less, proportionally, to the overall incidence.27 STD management thus remains a cornerstone of any program designed to help control HIV transmission3; however, appropriate coverage of endemic pathogens is often challenging.
Although no clinical trial has ever demonstrated that treating male trichomoniasis has resulted in a reduction of HIV transmission, epidemiologic data suggest that this pathogen is a transmission cofactor.2 Our data suggest that treating men for T. vaginalis when they first present to the STD clinic could be a cost-effective means to avert HIV infections. We estimate that increasing coverage to treat all men presenting with an STD (scheme C) could prevent between 6 and 52 new HIV infections per 10,000 men screened at costs ranging from $2 to $62 per infection averted. This compares favorably with other interventions in the developing world, such as blood screening, which costs $9.50 to $246 per case averted; voluntary counseling and testing, which costs $264 to $367 per case averted; and mother-to-child transmission interruption, which costs $83.70 to $5678 per case averted.28 A Tanzanian STD control program averted a case of HIV for approximately $260.11 Moses et al29 estimated the cost-effectiveness of STD treatment in a group of sex workers in Kenya; however, they used a high per-contact transmission estimate (1%) and also considered secondary transmission. Results from their sensitivity analysis using a scenario more consistent with ours (0.2% transmission, primary transmission only) suggest that the cost of averting a case of HIV ranged from $150 to $239.
Our cost-effectiveness analyses do not consider secondary or tertiary HIV infections resulting from primary HIV transmission attributable to T. vaginalis. We do not consider the transmission of T. vaginalis itself or the deleterious effects of untreated trichomoniasis. Additional visits to the clinic for unresolved STDs typically entail readministration of baseline (nonmetronidazole) therapy, a cost that we do not consider. Furthermore, our calculations only consider STD program costs from the government perspective and do not consider future medical costs related to HIV morbidity or other social costs to the client. These factors would only serve to tip the scales further in favor of adopting metronidazole.
Our data are taken from a single site, the largest STD clinic in the urban setting of Lilongwe. We believe that these data represent an accurate estimate of male trichomoniasis, because we used 2 detection assays, PCR and culture, and cultured the organism from 2 anatomic compartments, urine and the anterior urethra.30 Surveys of male attendees to other sub-Saharan African STD clinics have found this pathogen to be a common cause of NGU in men.5,6,16,31 T. vaginalis has been observed in 25% to 30% of women presenting for antenatal care in Blantyre in southern Malawi.32 We observed that nearly 13% of STD-asymptomatic men presenting to the dermatology outpatient clinic at the LCH were also infected.12 Working in neighboring Tanzania, Watson-Jones et al18 discovered that this pathogen was more than 7 times as prevalent as the next most common STD (11% vs. 1.5% for Chlamydia) and was associated with urethral inflammation to a similar degree as in cases of Neisseria gonorrhoeae and C. trachomatis. A survey of transport workers in Mombasa, Kenya also found similar results.33
Given that Malawi is in the mature, high-prevalence, high-transmission stage of its HIV epidemic, the population-level risk of HIV transmission attributable to STDs may be modest.27 Indeed, the number of cases averted that we report here is small enough that the sample size requirements to see an effect would proscribe conducting a randomized trial to evaluate treating trichomoniasis as a population-based intervention to reduce HIV transmission. We also observed that although metronidazole cleared infection with T. vaginalis and may have reduced the severity of urethral inflammation, it did not affect overall rates of urethral inflammation by 7 days after treatment.7
Fluctuations in the risk of HIV transmission attributable to T. vaginalis infection led to the largest changes to our ICERs (see Table 4). Estimating this attributable risk remains a significant hurdle to providing an accurate estimate of this intervention's cost-effectiveness. Although the published values for the increased risk of male-to-female HIV transmission attributable to trichomoniasis are all positive, the point estimates vary and are imprecise. Estimates of male susceptibility and infectiousness are based on the symptoms of urethritis and not specifically trichomoniasis. Our estimates for the increased risk of HIV transmission are drawn from a number of sources and compare favorably with a probabilistic model of HIV transmission.34 From data based on a limited number of men with trichomonal urethritis and seminal plasma HIV RNA values,7,8 we can use this model to calculate a per-contact transmission probability attributable to T. vaginalis at 0.007. Ten unprotected sexual contacts while infected with T. vaginalis approximates our transmission estimate. Recent data from a lower risk population in Uganda reports 36 contacts, on average, over the 4-month time period used in our study.35 Therefore, our transmission estimate may be conservative. The beneficial effects of treatment for trichomoniasis on HIV transmission (ie, reduction of urethral inflammation) may therefore take longer than 1 week to observe. Because of these issues, one must consider the ICERs here in aggregate as a potential range of costs and savings under the assumption that T. vaginalis affords a modest increase in the risk for HIV transmission and/or acquisition. Given this, our most extreme value for the ICER observed under the univariate analysis of plan C, $61.51 per case, averted under the extremely low transmission scenario (see Table 4) remains highly cost-effective.
Metronidazole is already in use in Malawi nationwide for treating women with symptoms of trichomoniasis and their asymptomatic partners. Therefore, the logistical challenges to expand treatment to men would be minimal. Additional training requirements would be negligible, especially if scheme C, universal treatment, were adopted. The analysis we conducted is unique in that it represents a simple add-on to an existing STD management scheme. No great capital outlay or recurrent expenditures beyond the cost of metronidazole and reprinting training materials are likely to be incurred above and beyond the current costs for the national STD management program.
Metronidazole is inexpensive, efficacious, and relatively well tolerated. It is easy to administer, and when it is given as a single dose, the likelihood of developing any clinically meaningful microbial resistance is low. Universal treatment of this pathogen in a clinic where the male prevalence is 20% adds up to $0.25 per case treated, which is less than the per-dose cost of all other drugs administered to men arriving for STD care. In this report, we show that adding metronidazole to an existing syndromic STD control program could be a highly cost-effective means to curb HIV transmission in this cohort. Even in the absence of an effect on HIV transmission, treatment of male trichomoniasis in this population of high-risk behavior symptomatic men represents an important opportunity to eradicate a reservoir of transmission to women. A serious rethinking of the current treatment protocols in Malawi and elsewhere may be in order.
1. UNAIDS. AIDS Epidemic Update: December 2005 Global Summary of the AIDS Epidemic
. Geneva, Switzerland; UNAIDS; 2005.
2. Røttingen J-A, Cameron DW, Garnett GP. A systematic review of the epidemiologic interactions between classic sexually transmitted diseases and HIV. How much is known? Sex Transm Dis
3. UNAIDS/WHO. Consultation on STD Interventions for Preventing HIV: What's the Evidence?
Geneva, Switzerland: UNAIDS/WHO; 2000.
4. Gerbase AC, Rowley JT, Heymann DH, et al. Global prevalence and incidence estimates of selected curable STDs. Sex Transm Infect
. 1998;74 (Suppl 1):S12-S16.
5. Pepin J, Sobela F, Deslandes S, et al. Etiology of urethral discharge in West Africa: the role of Mycoplasma genitalium and Trichomonas vaginalis. Bull World Health Organ
6. Morency P, Dubois MJ, Gresenguet G, et al. Aetiology of urethral discharge in Bangui, Central African Republic. Sex Transm Infect
7. Price M, Zimba D, Hoffman IF, et al. The addition of treatment for trichomoniasis to the syndromic management of urethritis in Malawi: a randomized clinical trial. Sex Transm Dis
8. Hobbs MM, Kazembe P, Reed AW, et al. Trichomonas vaginalis as a cause of urethritis in Malawian men. Sex Transm Dis
9. Van der Veen F, Fransen L. Drugs for STD management in developing countries: choice, procurement, cost, and financing. Sex Transm Infect
. 1998;74(Suppl 1):S166-S174.
10. Creese A, Floyd K, Alban A, et al. Cost-effectiveness of HIV/AIDS interventions in Africa: a systematic review of the evidence. Lancet
11. Gilson L, Mkanje R, Grosskurth H, et al. Cost-effectiveness of improved treatment services for sexually transmitted diseases in preventing HIV-1 infection in Mwanza Region, Tanzania. Lancet
12. Price MA, Miller WC, Kaydos-Daniels SC, et al. Trichomoniasis in men and HIV infection: data from two outpatient clinics at Lilongwe Central Hospital of Malawi. J Infect Dis
13. Kaydos SC, Swygard H, Wise SL, et al. Development and validation of a PCR-based enzyme-linked immunosorbent assay with urine for use in clinical research settings to detect Trichomonas vaginalis in women. J Clin Microbiol
14. Bowden FJ, Garnett GP. Trichomonas vaginalis epidemiology: parameterising and analysing a model of treatment interventions. Sex Transm Infect
15. Petrin D, Delgaty K, Bhatt R, et al. Clinical and microbiological aspects of Trichomonas vaginalis. Clin Microbiol Rev
16. Latif AS, Mason PR, Marowa E. Urethral trichomoniasis in men. Sex Transm Dis
17. Malawi National Statistics Office and ORC Macro. Malawi Demographic and Health Survey 2000
. Zomba, Malawi, and Calverton, MD: National Statistical Office and ORC Macro; 2001.
18. Watson-Jones D, Mugeye K, Mayaud P, et al. High prevalence of trichomoniasis in rural men in Mwanza, Tanzania: results from a population based study. Sex Transm Infect
19. Quinn TC, Wawer MJ, Sewankambo N, et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. Rakai Project Study Group. N Engl J Med
20. Fideli US, Allen SA, Musonda R, et al. Virologic and immunologic determinants of heterosexual transmission of human immunodeficiency virus type 1 in Africa. AIDS Res Hum Retroviruses
21. UNAIDS and Economic Commission for Africa. AIDS in Africa Country by Country
. Geneva, Switzerland: UNAIDS; 2000:131-135.
22. Anderson R. Transmission dynamics of sexually transmitted diseases. In: Holmes KK, et al., eds. Sexually Transmitted Diseases
. New York, NY: McGraw-Hill; 1999;25-37.
23. Fleming DT, Wasserheit JN. From epidemiological synergy to public health policy and practice: the contribution of other sexually transmitted diseases to sexual transmission of HIV infection. Sex Transm Infect
24. Picazo OF. Better Health Outcomes from Limited Resources: Focusing on Priority Services in Malawi
. World Bank, Washington, DC; 2002.
25. Ministry of Health and Population and National AIDS Commission. Management of Sexually Transmitted Infections Using Syndromic Management Approach: Guidelines for Service Providers
, Lilongwe, Malawi. 2003.
26. Taha TE, Canner JK, Chiphangwi JD, et al. Reported condom use is not associated with incidence of sexually transmitted diseases in Malawi. AIDS
27. Grosskurth H, Gray R, Hayes R, et al. Control of sexually transmitted diseases for HIV-1 prevention: understanding the implications of the Mwanza and Rakai trials. Lancet
28. Walker D. Cost and cost-effectiveness of HIV/AIDS prevention strategies in developing countries: is there an evidence base? Health Policy Plan
29. Moses S, Plummer FA, Ngugi EN, et al. Controlling HIV in Africa: effectiveness and cost of an intervention in a high-frequency STD transmitter core group. AIDS
30. Kaydos-Daniels SC, Miller WC, Hoffman I, et al. The use of specimens from various genitourinary sites in men to detect T. vaginalis infection. J Infect Dis
31. Pillay DG, Hoosen AA, Vezi B, et al. Diagnosis of Trichomonas vaginalis in male urethritis. Trop Geogr Med
32. Taha ET, Dallabetta GA, Hoover DR. Trends of HIV-1 and sexually transmitted diseases among pregnant and postpartum women in urban Malawi. AIDS
33. Jackson DJ, Rakwar JP, Cohan B, et al. Urethral infection in a workplace population of East African men: evaluation of strategies for screening and management. J Infect Dis
34. Chakraborty H, Sen PK, Helms RW, et al. Viral burden in genital secretions determines male-to-female sexual transmission of HIV-1: a probabilistic empiric model. AIDS
35. Gray RH, Wawer MJ, Brookmeyer R, et al. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV-1-discordant couples in Rakai, Uganda. Lancet
36. UNAIDS/WHO. Epidemiologic Fact Sheet on HIV/AIDS and Sexually Transmitted Diseases, Malawi
. Geneva, Switzerland: UNAIDS/WHO; 1998:12.
Keywords:© 2006 Lippincott Williams & Wilkins, Inc.
Africa; cost-effectiveness; HIV/AIDS; sexually transmitted diseases; syndromic STD management; T. vaginalis