SUMMARY
OBJECTIVE
To assess the cost-effectiveness of screening for latent tuberculosis infection (LTBI) using a commercially available detection test and treating individuals at high risk for human immunodeficiency virus (HIV) infection in a middle-income country.
DESIGN
We developed a Markov model to evaluate the cost per LTBI case detected, TB case averted and quality-adjusted life year (QALY) gained for a cohort of 1000 individuals at high risk for HIV infection over 20 years. Baseline model inputs for LTBI prevalence were obtained from published literature and cross-sectional data from tuberculosis (TB) screening using QuantiFERON®-TB Gold In-Tube (QFT-GIT) testing among sex workers and illicit drug users at high risk for HIV recruited through street outreach in Tijuana, Mexico. Costs are reported in 2007 US dollars. Future costs and QALYs were discounted at 3% per year. Sensitivity analyses were performed to evaluate model robustness.
RESULTS
Over 20 years, we estimate the program would prevent 78 cases of active TB and 55 TB-related deaths. The incremental cost per case of LTBI detected was US$730, cost per active TB averted was US$529 and cost per QALY gained was US$108.
CONCLUSIONS
In settings of endemic TB and escalating HIV incidence, targeting LTBI screening and treatment among high-risk groups may be highly cost-effective.
Keywords: cost-effectiveness, QuantiFERON, latent tuberculosis infection, screening
TUBERCULOSIS (TB) is a major threat to global public health which disproportionately affects low- and middle-income countries.1 The economic burden for treating new cases of TB surpasses one billion dollars per year worldwide,2 and in Mexico the National TB Program has increased spending by 7.5 times since 2001.3 Targeted screening and treatment for latent TB infection (LTBI) has been recognized as an effective strategy for TB control;4 however, it is not practiced in most middle- and low-income countries with a high TB burden.5 LTBI has traditionally been diagnosed using the tuberculin skin test (TST); however, the TST is limited by its lack of specificity, particularly in populations vaccinated with the bacille Calmette-Guérin (BCG).6
Mexico is a middle-income country with a moderate TB burden7 and a low prevalence of human immunodefi ciency virus (HIV) infection.8 However, the prevalence of both TB and HIV varies widely between regions within Mexico. States bordering the United States have alarming TB rates.9 The northwestern-most state of Baja California, which includes the border town of Tijuana, has the highest TB rate of all the border states (57.3 per 100 000 population)7 and the highest accumulated incidence of acquired immune-deficiency syndrome cases of all Mexican states (170/100 000), second only to Mexico City Federal District (240/100 000).8
High-risk groups for HIV infection, such as injection drug users (IDU), sex workers and homeless populations, are particularly vulnerable for progressing from LTBI to active TB,10,11 and their risk is 20 times higher than that of the general population.12
In Mexico, where most of the population has received the BCG vaccine as part of the standard immunization schedule, there are no programs that actively screen for TB or LTBI even in high-risk groups. The official Mexican guidelines recommend LTBI treatment only for exposed children aged <5 years, children between the ages of 5 and 14 years if they have no history or signs of BCG vaccination, and HIV-infected individuals with known exposure to an active case of TB.13
Until recently, the only test available for the detection of LTBI was the TST; however, recent advances have allowed the development of commercial in vitro T-cell-based interferon-gamma release assays (IGRA).14 These assays use antigens specific to Mycobacterium tuberculosis and are not affected by previous BCG vaccination, making it a useful screening test for LTBI in BCG-vaccinated populations such as Mexico.14 Published studies involving IGRAs to detect LTBI in high-income, low-prevalence countries have shown this intervention to be cost-effective.15 To date, no studies evaluating this approach in high-prevalence, limited-resource countries with BCG-vaccinated populations have been published.
Between March and August of 2007, our research team conducted a screening program for TB and LTBI targeted at adults with a high risk for HIV infection in a community-based setting in Tijuana (PreveTB Study). We obtained prevalence estimates of active TB, LTBI and HIV infection in 503 individuals with a high risk for HIV. The IGRA QuantiFERON®-TB Gold In-Tube (QFT-GIT), combined with a chest X-ray (CXR) and three sputum smears to detect acid-fast bacilli (AFB), was used to determine the presence of LTBI.16
The purpose of the analysis reported here is to assess the cost-effectiveness of LTBI screening and treatment in Tijuana among individuals at increased risk for HIV and TB, using data from the PreveTB study.
METHODS
We modeled the cost-effectiveness of a screening and treatment program for LTBI among 1000 adult individuals from Tijuana, Mexico, with a high risk for HIV infection. Current values for HIV and TB prevalence were derived from the PreveTB Study. Literature review and personal communications were used to inform other model inputs.
The cost-effectiveness analysis was approved as part of the PreveTB Study by the Institutional Review Board of the University of California, San Diego, and the Ethics Board of the Tijuana General Hospital.
Targeted LTBI screening
The model portrays a community-based program in an urban, quasi-legal commercial sex district with a high concentration of sex workers, homeless individuals and illicit drug users.16 In this program, participants at high risk for HIV were given a short interview to investigate TB symptoms.17 Immediately after the initial interview, blood was drawn for QFT-GIT and HIV tests. Participants meeting specific symptom criteria suggestive of active TB were asked to provide a sputum sample for AFB smears and two additional samples on 2 consecutive days following the initial encounter. One week later, all subjects received test results for QFT-GIT and HIV tests. Subjects with positive HIV test results were referred to the municipal HIV clinic for medical care. Individuals with a positive QFT-GIT (but a negative smear) were offered a CXR. Subjects with a normal CXR were considered to have LTBI. Those with abnormal CXRs were referred to the municipal TB clinic and studied for active TB.
Hypothetical LTBI treatment
In the model, individuals diagnosed with LTBI infection would be offered treatment consisting of 6 months of 900 mg isoniazid (INH) twice weekly18 under community directly observed therapy (DOT). An incentive payment of US$40 per month was included in the model to increase adherence.11
Markov model
We constructed a Markov model using TreeAge Suite software (TreeAge Software Inc, V. 1.31, Williamstown, MA, USA). The cost of the program was defined as the sum of screening and LTBI treatments administered, multiplied by the unit costs for each of the tests and treatments, minus the savings in medical costs due to averted TB disease and mortality. The expected health benefit is calculated on the basis of the estimated baseline TB disease incidence in susceptible individuals and the estimated reduction in that incidence due to a screening and treatment program.
A seven-state transitional Markov model was used to determine the cost-effectiveness over a 20-year period. In each 1-year cycle, individuals can remain in one of the seven mutually exclusive states or transit between them. State 1 is no LTBI or HIV. The next two states involve TB and no HIV: LTBI (State 2); active TB disease (State 3). Individuals can progress from any state to death (State 7) or be cured of TB (State 1). Individuals can transition into three HIV states: HIV infection without TB (State 4); HIV and LTBI co-infection (State 5); and active TB and HIV co-infection (State 6).
Epidemiological data inputs
Baseline epidemiological parameters that characterize the population include the prevalence of LTBI, active TB and HIV infection in the target population (Table 1). Epidemiological parameters that characterize the state transitional probabilities within the model are reported in Table 2: the annual risk for LTBI and HIV infection, protective effect against TB morbidity of 6 months of INH for LTBI, and increased adherence due to financial incentives.11 The protective effect of 6 months of INH treatment for LTBI infection was calculated at 69%, the protection period was assumed to be 5 years to account for the possibility of re-infection,21 and baseline utility weight for living with active TB was calculated at 0.58.27,28 Estimates on QFT-GIT sensitivity and specifi city for LTBI and active TB detection were based on extensive review of the literature.6,15,23
Table 1.
Initial distribution between health states
| Health state | % (range) |
|---|---|
| 1 No TB/no LTBI/no HIV | 39.5 (variable*) |
| 2 LTBI/no HIV | 56 (45–67) |
| 3 Active TB/no HIV | 0.2 (0.2–4) |
| 4 HIV/no TB | 1.8 (0.15–3) |
| 5 HIV+LTBI† | 2.3 (0.15–3) |
| 6 HIV+active TB‡ | 0.2 (0.2–1) |
| 7 Death | — |
Value changes to sum 1 for sensitivity analyses.
HIV and LTBI co-infection.
HIV and active TB co-infection.
TB = tuberculosis; LTBI = latent TB infection; HIV = human immunodeficiency virus.
Table 2.
Inputs for model assessing the cost-effectiveness of a screening and treatment program for latent tuberculosis infection in Tijuana, Mexico
| Input category / input | Base-case value | Range | Source |
|---|---|---|---|
| Costs, US$ | |||
| QFT-GIT test kit | 19.00 | 15.00–45.00 | *15 |
| AFB smears in sputum | 5.50 | 3.50–7.50 | *19 |
| Chest X ray | 25.00 | 10.00–35.00 | *19 |
| INH for 6 months† | 2.16 | 2.16–44.00 | 20 |
| Hepatitis due to INH toxicity | 250 | 150–750 | 19 |
| Incentives for 6 months | 240 | 120–360 | 11 |
| HIV rapid tests (determine) | 4.00 | — | *20 |
| HIV confirmatory test (IFA) | 50.00 | 4.59–50‡ | * |
| Fixed costs running program/year§ | 150 | 100–300 | * |
| Start up costs (year 1)§ | 50 | 30–100 | * |
| Active TB management | 10 342 | 5000–15 000 | 19 |
| Effectiveness | |||
| Protective effect of 6 months of | |||
| INH treatment¶ | 0.69 | 0.21–0.75 | 4,21,22 |
| Adherence to LTBI treatment | 0.80 | 0.30–0.85 | 4,10,11 |
| Protective effect of 6 months of | |||
| INH treatment | 5 yrs | 3–20 years | 22 |
| Sensitivity of QFT-GIT | 0.95 | 0.85–1 | 6,15,23 |
| Specificity of QFT-GIT | 0.98 | 0.85–1 | 6,15,23 |
| Transitions | |||
| Annual HIV risk# | 2 | 0.3–4 | 24 |
| Annual LTBI risk# | 5 | 1–20 | 5 |
| Annual risk of progression from | |||
| LTBI to active TB# | 1.1/5.4** | 0.065–5.4 | 10 |
| Probability of death from active TB | |||
| without treatment# | 16.5/35** | — | 10 |
| Probability of INH toxicity# | 18 | 10–30 | 25 |
| Probability of death from other | |||
| causes (age 30) | Adjusted life tables | — | 26 |
PreveTB study costs.
Administrative and delivery expenses are included in the fixed program costs.
Alternative use of Unigold rapid HIV test (Trinity Biotech Inc. St Louis, MO, USA) for confirmation.
Cost calculated per participant.
For 80% adherence to LTBI treatment.
Per 100 person-years.
Value used for HIV-positive individuals.
QFT-GIT = QuantiFERON-TB Gold In-Tube; AFB = acid-fast bacilli; INH = isoniazid; HIV = human immunodeficiency virus; IFA = immunofluorescence assay; TB = tuberculosis; LTBI = latent TB infection.
Costs
Our cost-effectiveness analysis adopted a government health care payer perspective as deemed relevant to inform public policy decisions in Mexico and other low-to middle-income countries.29 Costs are reported in 2007 US dollars, and future costs and quality-adjusted life years (QALYs) are discounted at 3% annually, as recommended by the US Panel on Cost-Effectiveness in Health and Medicine and other analysts.30
Specific cost inputs were obtained using an ingredients approach for the observed cost per screening during the cross-sectional PreveTB study. A step-down approach was used for fixed costs such as space rental and personnel expenditures.31 The costs associated with adverse events due to LTBI treatment and the future costs for managing active TB disease were obtained from published reports. Costs are summarized in Table 2.
Outcomes
The model calculated the number of LTBI cases identified, TB cases averted, TB-related deaths averted, and QALYs gained over a 20-year period comparing the hypothetical targeted LTBI screening program vs. no program.
The economic outcomes for the analysis were incremental costs per LTBI case detected, per active TB case averted, per TB-related death averted, and per QALY gained. Based on the gross domestic product (GDP) per capita in Mexico, a threshold of $10 000 was used for willingness to pay for additional QALY gained.32
Sensitivity analyses
We conducted sensitivity analyses to assess the effect of uncertainty in all input values according to the 95% confidence intervals (CIs) derived from the PreveTB Study and for the likely range of other inputs to encompass low and high values (Table 2). We used one-way sensitivity analyses for all inputs and reported those with substantial effects on results. We also used two-way sensitivity analyses to explore the effects on the model due to changes in epidemiologic variables.
We used Monte Carlo simulation to conduct a probabilistic sensitivity analysis and determine 95% uncertainty ranges. This method33 involves 100 000 individual random walks through the model simulated one at a time. Each walk is assigned to one of the starting health states by the software using tables created from the data obtained from the PreveTB study.
RESULTS
Base case results
Base case results are summarized in Table 3. For a population with a high risk for HIV infection, the incremental cost-effectiveness ratio (ICER) for detecting a case of LTBI is US$730, for preventing a case of active TB it is US$529 and for preventing a TB-related death it is US$737. The cost for each additional QALY gained was US$108. The total cost for the LTBI screening and treatment program, excluding costs related to treatment of active TB, was US$433 000 for Year One and US$380 000 for subsequent years. In 20 years, the LTBI screening and treatment program would result in a 22% reduction in expected spending for managing active TB cases compared with the absence of the LTBI program (US$9.13 million vs. US$11.64 million = US$2.51 million savings).
Table 3.
Base-case results (high-risk population for HIV and TB)
| No | LTBI | ||
|---|---|---|---|
| Outcomes | program | program | Difference |
| Number of active TB cases* | 255 | 177 | 78 |
| Number of TB-related deaths* | 80 | 25 | 55 |
| QALYs* | 11612 | 11992 | 380 |
| Incremental cost per QALY | |||
| gained, US$ | — | — | 108 |
| Incremental cost per LTBI case | |||
| detected, US$ | — | — | 730 |
| Incremental cost per active TB | |||
| case prevented, US$ | — | — | 529 |
| Incremental cost per TB death | |||
| averted, US$ | — | — | 737 |
Based on 1000 high-risk individuals for HIV and TB.
HIV = human immunodeficiency virus; TB = tuberculosis; QALY = quality-adjusted life year; LTBI = latent TB infection.
Sensitivity analyses
Results for one-way sensitivity analyses are shown in Table 4. Sensitivity analysis varying costs for treating active TB cases changed the cost-effectiveness from US$650 per QALY gained at a cost of US$5000 for managing active TB cases to net savings at average costs for managing active TB cases of above US$12 000. Changes in HIV prevalence showed improved cost-effectiveness, with an incremental cost per QALY gained ranging from US$253 at an HIV prevalence of 0.3% to net savings of US$107 at an HIV prevalence of above 6%. In contrast, changes in active TB prevalence had the opposite effect on cost-effectiveness, with a higher marginal cost per QALY gained at a higher baseline prevalence of active TB; this is due mainly to the lower number of individuals with LTBI who can benefit from INH preventive therapy. None of the scenarios gave results above the US$10 000 per QALY gained threshold for willingness to pay.
Table 4.
Results of one-way sensitivity analysis using plausible ranges
| Variable | Range | Cost-effectiveness |
|---|---|---|
| QFT-GIT test kit costs, US$ | 5–45 | 70–180 |
| Cost for chest X rays, US$ | 5–50 | 82–125 |
| Cost for incentives, US$* | 20–60 | Net savings†–260 |
| HAART coverage, % | 0–100 | 108–780 |
| Annual discounting, % | 2–10 | 21–1162 |
| Cost for treating active TB, US$ | 500–15 000 | 1050–net savings† |
| QFT-GIT sensitivity, % | 85–100 | 160–88 |
| QFT-GIT specificity, % | 85–100 | 170–100 |
| Annual HIV incidence‡ | 0.3–4 | 708–net savings† |
| Annual TB incidence‡ | 0.065–2 | 178–80 |
| Years of INH protection | 3–10 | 180–47 |
| Operational costs, US$§ | 100–300 | Net savings†–250 |
| Cost for TB case treatment, US$ | 5000–12 000 | 641–net savings† |
| Annual risk of LTBI for | ||
| HIV-infected individuals§ | 5–25 | 108–80 |
Per month while under 6 months of INH treatment.
Net savings is reached when program saves money for additional QALY gained.
Per 100 person-years.
For sensitivity analysis, costs are calculated per participant.
QFT-GIT = QuantiFERON-TB Gold In-Tube; HAART = highly active antiretroviral therapy; TB = tuberculosis; HIV = human immunodeficiency virus; INH = isoniazid; LTBI = latent TB infection.
Changes reflecting different scenarios are shown in Table 5. The model was sensitive to changes in the incidence of TB and HIV. When outcomes are measured over a 20-year time horizon, the ICER for the LTBI screening and treatment program remains below US$1000 for additional QALYs in all scenarios, except when we consider HIV and TB risks reflective of the general population in Tijuana, increasing to US$8352 per QALY gained, US$36 748 for each additional active TB case averted and US$45 935 for each additional TB death averted. A two-way sensitivity analysis for changes in annual risk of HIV infection and average cost of managing active TB cases showed that the LTBI program becomes cost-saving when the annual risk of HIV is above 6 per 100 person-years at costs for managing active TB of US$6000 or more, and when the HIV annual risk is 0.3 per 100 person-years at costs for managing active TB of above US$15 000 (Figure).
Table 5.
Sensitivity analysis for changes in HIV and TB risk modeled for 1000 individuals
| Incremental | Incremental | |||||||
|---|---|---|---|---|---|---|---|---|
| Total | Incremental | Cost/TB case | Cost/averted | |||||
| Total cost | Total | Total | TB | Cost/QALY | Prevented | TB death | ||
| Scenario | Strategy | US$ | QALYs | Cases | deaths | US$ | US$ | US$ |
| Lower HIV risk | No program | 1 173 328 | 12 574 | 220 | 76 | — | — | — |
| HIV risk: 0.3/100 py | LTBI program | 1 427 196 | 12 932 | 152 | 18 | 709 | 3733 | 4 377 |
| TB risk: 1.1/100 py | Difference | 253 868 | 358 | 68 | 58 | |||
| Moderate HIV risk | No program | 1 362 942 | 12 158 | 236 | 78 | |||
| HIV risk: 1/100 py | LTBI program | 1 522 872 | 12 526 | 162 | 20 | 435 | 2 161 | 2 757 |
| TB risk: 1.1/100 py | Difference | 159 930 | 368 | 74 | 58 | |||
| High HIV risk | No program | 1 991 310 | 10 672 | 284 | 86 | |||
| HIV risk: 4/100 py | LTBI program | 1 839 706 | 11 068 | 200 | 30 | Net savings* | Net savings* | Net savings* |
| TB risk: 1.1/100 py | Difference | Net savings | 396 | 84 | 56 | |||
| Lower TB risk | No program | 893 764 | 11 956 | 104 | 28 | |||
| HIV risk: 2/100 py | LTBI program | 965 778 | 12 090 | 68 | 12 | 537 | 2 000 | 4 501 |
| TB risk: 0.065/100 py | Difference | 72 014 | 134 | 36 | 16 | |||
| Moderate HIV and TB risk | No program | 994 879 | 12 338 | 157 | 50 | |||
| HIV risk: 1/100 py | LTBI program | 1 172 333 | 12 576 | 106 | 15 | 744 | 3 479 | 5 070 |
| TB risk: 0.6/100 py | Difference | 177 454 | 238 | 51 | 35 | |||
| General population in Tijuana | No program | 230 188 | 13 174 | 26 | 12 | |||
| HIV risk: .0073/100 py | LTBI program | 597 668 | 13 218 | 16 | 4 | 835 | 36 748 | 45 935 |
| TB risk: 0.065/100 py | Difference | 367 480 | 44 | 10 | 8 |
Net savings is reached when program saves money for active TB cases prevented compared with expenditures for screening and treatment of LTBI.
HIV = human immunodeficiency virus; TB = tuberculosis; QALY = quality-adjusted life year; py = person-years.
Figure.
Two-way sensitivity analysis examining the impact of annual risk of HIV and average cost for managing active TB cases on potential cost savings from the LTBI screening and treatment program used in the model. TB = tuberculosis; HIV = human immunodeficiency virus; LTBI = latent TB infection.
We compared the cost-effectiveness of the LTBI screening and treatment program between the base-case scenario and a population with lower annual HIV and TB risk, similar to the general population of Tijuana, Mexico, without considering expenditures from managing active TB cases (and thus savings from averted active TB). The cost in the base-case scenario was US$1143 per QALY gained compared with US$10 150 per QALY gained in the general population. Table 6 shows the cost per QALY gained, TB cases prevented and TB deaths averted while comparing cost-effectiveness between individuals at high risk for HIV-TB and for a population with lower risk for HIV and TB.
Table 6.
Cost-effectiveness of the LTBI screening and treatment program, excluding costs of managing active TB and comparing a population with high risk for HIV and TB to a population with TB and HIV risks similar to the general population in Tijuana, Mexico
| Incremental cost | |
|---|---|
| Outcomes | per occurrence, US$ |
| High-risk groups for HIV-TB* | |
| Active TB case prevented | 5 000 |
| TB death averted | 7 500 |
| QALY gained | 1 143 |
| General population in Tijuana† | |
| Active TB case prevented | 40 000 |
| TB death averted | 69 400 |
| QALY gained | 10 150 |
HIV risk: 2/100 py; TB risk: 1.1/100 py.
HIV risk: 0.0073/100 py; TB risk: 0.065/100 py.
LTBI = latent TB infection; TB = tuberculosis; HIV = human immunodeficiency virus; QALY = quality-adjusted life year; py = person-years.
The Monte Carlo simulations for 100 000 trials resulted in a range for cost-effectiveness between cost savings (US$200 saved) to US$412 per additional QALY gained by a program targeted to individuals at risk for HIV.
DISCUSSION
This analysis shows that, in a middle-income country with high TB prevalence and escalating HIV prevalence, the use of QFT-GIT for screening of LTBI could have a public health benefit and be cost-effective by reasonable standards. In fact, the TB screening and treatment program considered in this analysis was more cost-effective (US$108/QALY) than other accepted preventive interventions such as routine Papanicolaou test for cervical cancer in sexually active women (>US$14 000/QALY), colorectal screening in older adults using fecal occult blood test (US$10 000/QALY), and routine diabetes screening in adults and diet counseling for patients at risk for cardiovascular disease (>US$14 000/QALY).34 Compared to other accepted TB control interventions in resource-limited settings, this program can be as cost-effective as TB contact tracing in adults and only slightly more expensive than DOT for active TB (US$91/QALY).1
Based on World Health Organization (WHO) recommendations,32 we used the per capita gross domestic product in Mexico as a threshold to evaluate cost-effectiveness, which showed that this program is cost-effective for all scenarios, except when we modeled TB and HIV incidence rates similar to the general population of Tijuana and averted costs for future active TB cases were not included. Further research may be needed to determine relevant thresholds for willingness to pay for additional QALYs gained from an intervention based on the perspective of Mexican health officials. Our findings largely depend on the effectiveness of program performance. For example, even when DOTS has been proven to be one of the most cost-effective interventions in TB control,1 reports from Baja California, Mexico, have raised serious questions about the effectiveness of DOT.35 LTBI screening is useless unless future cases of active TB are in fact prevented. Therefore, the decision to test must include a well-planned and effective treatment program, regardless of the screening method.
Certain limitations must be considered. There is currently no gold standard for LTBI diagnosis, and uncertainty regarding true LTBI prevalence remains. As the PreveTB study did not include treatment or long-term follow-up, outcome parameters to evaluate effectiveness (e.g., treatment efficacy, TB-related illness and mortality) used in our model were based on published reports that may not accurately reflect our study population. Both limitations support the urgent need for longitudinal studies in Mexico to further elucidate the effectiveness of LTBI screening and treatment interventions. As additional research is published, model inputs can be updated, thus narrowing the ranges needed for sensitivity analyses. In conclusion, our study suggests that within a population at high risk for HIV and TB, targeted LTBI screening and treatment programs using QFT-GIT can be highly cost-effective.
Acknowledgments
The PreveTB study was funded by the United States Agency for International Development (USAID) grant number: GSM-025. USAID TIES grant number: 523-A-00-06-00009-00. The El Cuete Study was funded by NIDA grant number: R01 DA019829. JLB was supported by a NIH T32 training grant: A107384 NIDA Diversity Supplement grant number: R01 DA023877-S1. Quanti-FERON®-TB Gold In-Tube assay training provided by Cellestis Inc (Carnegie, VIC, Australia).
References
- 1.Baltussen R, Floyd K, Dye C. Achieving the millennium development goals for health: cost effectiveness analysis of strategies for tuberculosis control in developing countries. BMJ. 2005;331:1364–1368. doi: 10.1136/bmj.38645.660093.68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Floyd K, Blanc L, Raviglione M, Lee JW. Resources required for global tuberculosis control. Science. 2002;295:2040–2041. doi: 10.1126/science.1069771. [DOI] [PubMed] [Google Scholar]
- 3.Castellanos M. Health system reform and political will: do these interventions enable reduction in TB burden?. 12th Annual Conference of the International Union Against Tuberculosis and Lung Disease-North American Region; 29 February 2008; San Diego, CA, USA. Vancouver, BC, Canada: British Columbia Lung Association; 2008. [Accessed May 2009]. http://www.bc.lung.ca/lungdiseases/iuatld-pdfs/2008/Health_System_Reform_and_Political_Will-Do_These_Interventions_Enable_Reduction_in_TB_Burden.pdf. [Google Scholar]
- 4.Gourevitch MN, Alcabes P, Wasserman WC, Arno PS. Cost-effectiveness of directly observed chemoprophylaxis of tuberculosis among drug users at high risk for tuberculosis. Int J Tuberc Lung Dis. 1998;2:531–540. [PubMed] [Google Scholar]
- 5.Bell JC, Rose DN, Sacks HS. Tuberculosis preventive therapy for HIV-infected people in sub-Saharan Africa is cost-effective. AIDS. 1999;13:1549–1556. doi: 10.1097/00002030-199908200-00016. [DOI] [PubMed] [Google Scholar]
- 6.Harada N, Nakajima Y, Higuchi K, Sekiya Y, Rothel J, Mori T. Screening for tuberculosis infection using whole-blood interferon-gamma and Mantoux testing among Japanese healthcare workers. Infect Control Hosp Epidemiol. 2006;27:442–448. doi: 10.1086/504358. [DOI] [PubMed] [Google Scholar]
- 7.Secretaria de Salud. Salud Mexico. Mexico City, Mexico: Direccion General de Evaluacion y Desempeño; 2004. 2005. [Google Scholar]
- 8.Centro Nacional para la Prevención y el Control del VIH/ SIDA. Consejo Nacional para la Prevencion y Control del VIH/ SIDA (CONASIDA). AIDS by numbers. Mexico City, Mexico: CENSIDA; 2006. [Accessed March 2008]. [Spanish] http://www.salud.gob.mx/conasida/estadis/2006/elsidaencifras15nov2006.pdf. [Google Scholar]
- 9.Tuberculosis Along the US-Mexico Border Work Group. Centers for Disease Control and Prevention. Preventing and controlling tuberculosis along the US-Mexico border. MMWR. 2001;50(RR1):1–27. [PubMed] [Google Scholar]
- 10.Snyder DC, Paz EA, Mohle-Boetani JC, Fallstad R, Black RL, Chin DP. Tuberculosis prevention in methadone maintenance clinics. Effectiveness and cost-effectiveness. Am J Respir Crit Care Med. 1999;160:178–185. doi: 10.1164/ajrccm.160.1.9810082. [DOI] [PubMed] [Google Scholar]
- 11.Tulsky JP, Hahn JA, Long HL, et al. Can the poor adhere? Incentives for adherence to TB prevention in homeless adults. Int J Tuberc Lung Dis. 2004;8:83–91. [PubMed] [Google Scholar]
- 12.Figueroa-Munoz J, Ramon-Pardo P. Tuberculosis control in vulnerable groups. Bull World Health Organ. 2008;86:657–736. doi: 10.2471/BLT.06.038737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.ISESALUD. Modificacion de la Norma Oficial Mexicana NOM-006-SSA-1993. Mexico City, Mexico: ISESALUD; 1993. pp. 3–10. [Google Scholar]
- 14.Diel R, Nienhaus A, Lange C, Meywald-Walter K, Forssbohm M, Schaberg T. Tuberculosis contact investigation with a new, specific blood test in a low-incidence population containing a high proportion of BCG-vaccinated persons. Respir Res. 2006;7:77. doi: 10.1186/1465-9921-7-77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Oxlade O, Schwartzman K, Menzies D. Interferon-gamma release assays and TB screening in high-income countries: a cost-effectiveness analysis. Int J Tuberc Lung Dis. 2007;11:16–26. [PubMed] [Google Scholar]
- 16.Garfein RS. Screening and diagnosis for TB among populations at risk for HIV infection in Tijuana, Mexico. 12th Annual Conference of the International Union Against Tuberculosis and Lung Disease-North American Region; 29 February 2008; San Diego, CA, USA. Vancouver, BC, Canada: British Columbia Lung Association; 2008. [Accessed May 2009]. http://www.bc.lung.ca/lungdiseases/iuatld-pdfs/2008/PREVE_TB_USAID-UC_San_Diego.pdf. [Google Scholar]
- 17.Catanzaro A, Perry S, Clarridge JE, et al. The role of clinical suspicion in evaluating a new diagnostic test for active tuberculosis: results of a multicenter prospective trial. JAMA. 2000;283:639–645. doi: 10.1001/jama.283.5.639. [DOI] [PubMed] [Google Scholar]
- 18.American Thoracic Society. Targeted tuberculin testing and treatment of latent tuberculosis infection. MMWR. 2000;49(RR6):1–51. [PubMed] [Google Scholar]
- 19.Vargas M, Rios L, Salazar MA, Cano F. Costos de atencion de la tuberculosis: caso del Instituto Nacional de Enfermedades Respiratorias (INER) Rev Inst Nal Enf Resp Mex. 2003;16:219–225. [Spanish] [Google Scholar]
- 20.IDA Foundation. Price indicator. Amsterdam, The Netherlands: IDA Foundation; 2007. [Google Scholar]
- 21.Comstock GW. How much isoniazid is needed for prevention of tuberculosis among immunocompetent adults? [Counterpoint] Int J Tuberc Lung Dis. 1999;3:847–850. [PubMed] [Google Scholar]
- 22.Efficacy of various durations of isoniazid preventive therapy for tuberculosis 5 years of follow-up in the International Union Against Tuberculosis trial. Bull World Health Organ. 1982;60:555–564. [PMC free article] [PubMed] [Google Scholar]
- 23.Mori T, Sakatani M, Yamagishi F, et al. Specific detection of tuberculosis infection: an interferon-γ-based assay using new antigens. Am J Respir Crit Care Med. 2004;170:59–64. doi: 10.1164/rccm.200402-179OC. [DOI] [PubMed] [Google Scholar]
- 24.Strathdee SA, Magis-Rodriguez C. Mexico’s evolving HIV epidemic. JAMA. 2008;300:571–573. doi: 10.1001/jama.300.5.571. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Nolan CM, Goldberg SV, Buskin SE. Hepatotoxicity associated with isoniazid preventive therapy: a 7-year survey from a public health tuberculosis clinic. JAMA. 1999;281:1014–1018. doi: 10.1001/jama.281.11.1014. [DOI] [PubMed] [Google Scholar]
- 26.Vlahov D, Wang C, Ompad D, et al. Mortality risk among recent-onset injection drug users in five US cities. Subst Use Misuse. 2008;43:413–428. doi: 10.1080/10826080701203013. [DOI] [PubMed] [Google Scholar]
- 27.Mahaparra P, Salomon JA, Nanda L. Measuring health state values in developing countries—results from a community survey in Andhra Pradesh. In: Murray CJ, Salomon JA, Murray JA, et al., editors. Summary measures of population health: concepts, ethics, measurement and applications. Geneva Switzerland: World Health Organization; 2002. pp. 473–485. [Google Scholar]
- 28.Dye C, Scheele S, Dolin P, Pathania V, et al. Global burden of tuberculosis: estimated incidence, prevalence and mortality by country. JAMA. 1999;282:677–686. doi: 10.1001/jama.282.7.677. [DOI] [PubMed] [Google Scholar]
- 29.Scott G. Towards a consistent payer perspective: a microeconomic approach to economic evaluation; Annual meeting of the Economics of Population Health: Inaugural Conference of the American Society of Health Economists; 4 June 2006; Madison, WI, USA: All Academic; 2006. [Google Scholar]
- 30.Gold MR. Standardizing cost-effectiveness analyses: the panel on cost-effectiveness in health and medicine. Acad Radiol. 1998;5 Suppl 2:S351–S354. doi: 10.1016/s1076-6332(98)80352-8. [DOI] [PubMed] [Google Scholar]
- 31.Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL. Methods for the economic evaluation of health care programmes. Oxford, UK: Oxford University Press; 2005. [Google Scholar]
- 32.Baltussen R, Adam T, Torres TT. Making choices in health. [Accessed October 2008];Geneva, Switzerland: World Health Organization; WHO guide to cost-effectiveness analysis. 2003 http://uclibs.org/PID/130450/10062367.
- 33.Briggs A, Sculpher M. An introduction to Markov modelling for economic evaluation. Pharmacoeconomics. 1998;13:397–409. doi: 10.2165/00019053-199813040-00003. [DOI] [PubMed] [Google Scholar]
- 34.Michael VM, Ashley BC, Nichol ME, Thomas JF, Michael JG, Leif IS. Priorities among effective clinical preventive services: results of a systematic review and analysis. Am J Prev Med. 2006;31:52–61. doi: 10.1016/j.amepre.2006.03.012. [DOI] [PubMed] [Google Scholar]
- 35.Radilla-Chavez P, Laniado-Laborin R. Results of directly observed treatment for tuberculosis in Ensenada, Mexico: not all DOTS programs are created equally. Int J Tuberc Lung Dis. 2007;11:289–292. [PubMed] [Google Scholar]