- Email: [email protected]
Cost effectiveness of high resolution computed tomography with interferon-gamma release assay for tuberculosis contact investigation
European Journal of Radiology 82 (2013) 1353–1358
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Cost effectiveness of high resolution computed tomography with interferon-gamma release assay for tuberculosis contact investigation Akiko Kowada ∗ Kojiya Haneda Healthcare Service, Ota City Public Health Office, Tokyo, Japan
a r t i c l e
i n f o
Article history: Received 30 October 2012 Received in revised form 2 February 2013 Accepted 9 February 2013 Keywords: High resolution computed tomography Interferon-gamma release assay Cost-effectiveness Tuberculosis contacts Chest X-ray
a b s t r a c t Background: Tuberculosis contact investigation is one of the important public health strategies to control tuberculosis worldwide. Recently, high resolution computed tomography (HRCT) has been reported as a more accurate radiological method with higher sensitivity and specificity than chest X-ray (CXR) to detect active tuberculosis. In this study, we assessed the cost effectiveness of HRCT compared to CXR in combination with QuantiFERON® -TB Gold In-Tube (QFT) or the tuberculin skin test (TST) for tuberculosis contact investigation. Methods: We constructed Markov models using a societal perspective on the lifetime horizon. The target population was a hypothetical cohort of immunocompetent 20-year-old contacts with smear-positive tuberculosis patients in developed countries. Six strategies; QFT followed by CXR, QFT followed by HRCT, TST followed by CXR, TST followed by HRCT, CXR alone and HRCT alone were modeled. All costs and clinical benefits were discounted at a fixed annual rate of 3%. Results: In the base-case analysis, QFT followed by HRCT strategy yielded the greatest benefit at the lowest cost ($US 6308.65; 27.56045 quality-adjusted life-years [QALYs])[year 2012 values]. Cost-effectiveness was sensitive to BCG vaccination rate. Conclusions: The QFT followed by HRCT strategy yielded the greatest benefits at the lowest cost. HRCT chest imaging, instead of CXR, is recommended as a cost effective addition to the evaluation and management of tuberculosis contacts in public health policy. © 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Tuberculosis (TB) is a widespread infectious and serious disease for global public health. To control TB, TB contact investigations play a very important role in public health policies not only by diagnosing latent tuberculosis infection (LTBI) and allowing chemoprophylaxis but also by earlier detection of active TB, often leading to treatment before appearance of symptoms of TB. The main approach for detecting active pulmonary TB is currently radiological examination by chest X-ray (CXR). However, low sensitivity and specificity of CXR are well known limitations. Recently, chest computed tomography has been reported as a more accurate and effective radiological method with higher sensitivity and specificity [1–4]. Multi-detector row computed tomography (MDCT) has greatly increased the clinical indications for CT and promotes to minimize excessive CT radiation exposure. High resolution CT
∗ Tel.: +81 3 3743 8315; fax: +81 3 3742 3116. E-mail address: [email protected] 0720-048X/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejrad.2013.02.017
(HRCT) images can be created from raw-data of MDCT without additional radiation [5–7]. Mycobacterium tuberculosis-specific interferon-gamma release assays (IGRAs) – QuantiFERON® TB Gold In-Tube (QFT) [Cellestis Limited, Chadstone, VIC, Australia] and T-SPOT® .TB (Oxford Immunotec, Oxford, UK) are now available and provide more accurate and sensitive diagnosis of M. tuberculosis infection with higher specificity than that of the tuberculin skin test (TST). They are not affected by bacillus Calmette-Guérin (BCG) vaccination and do not suffer from the booster phenomenon seen with repeated TSTs. However, these tests do not discriminate between active TB and LTBI and thus those testing positive are usually screened for active pulmonary TB using CXR. There are recent reports demonstrating the utility of using HRCT, rather than CXR, in combination with IGRAs for tuberculosis screening [7]. However, although more accurate, the purchase costs of HRCT and IGRAs are higher than CXR and TST and thus their overall cost effectiveness as mass TB screening tools warrants evaluation. In this study, we assessed the cost effectiveness of HRCT versus CXR in combination with QFT or TST strategies for TB contact investigation to demonstrate the optimal screening method for contacts.
1354
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
2. Methods 2.1. Target population Immunocompetent 20-year-olds contacts were chosen as a hypothetical cohort on lifetime horizon in developed countries. 2.2. Markov models The following four clinical states were included in our model to represent the possible clinical states in the target population: (i) Well (no LTBI and no TB); (ii) LTBI; (iii) TB; (iv) Death. Decision-analytical calculations were performed using Tree Age Pro Healthcare Module 2009 (Tree Age Software Inc., Williamstown, MA, USA). Each cycle length was 1 year. As this was a modeling study with all inputs and parameters derived from published literatures, ethics approval was not required. Markov models were developed for six strategies: QFT followed by CXR, QFT followed by HRCT, TST followed by CXR, TST followed by HRCT, CXR alone and HRCT alone (Fig. 1). The contacts were stratified by BCG-vaccination status for TST strategies. Markov models are state transition models to calculate expected values, costs and utilities [8]. (1) QFT followed by CXR strategy: The contact undergoes QFT testing. If QFT is positive, active TB is detected on CXR, and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If QFT is positive and active TB is not detected on CXR, the contact is treated per the standard 9-month isonicotinyl hydrazide (INH) chemoprophylaxis regimen for LTBI. If QFT is negative, the contact has no CXR and no need for follow-up. We considered the adherence and complication rates of chemoprevention [9,10]. We also estimated the recurrence of TB [11]. We used published estimates of sensitivity and specificity of QFT from a meta-analysis of developed country studies [12]. We used published estimates of sensitivity and specificity of CXR [13]. (2) QFT followed by HRCT strategy: The contact undergoes QFT testing. If QFT is positive, active TB is detected on HRCT, and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on HRCT, the contact is treated per the standard 9-month INH chemoprophylaxis for LTBI. If QFT is negative, the contact has no HRCT and no need for follow-up. We used published estimates of sensitivity and specificity of HRCT [1–4]. (3) TST followed by CXR strategy: The contact undergoes TST testing. If TST induration diameter is ≥5 mm in non BCG-vaccinated contacts and ≥10 mm in BCG-vaccinated contacts, the contact undergoes CXR. If active TB is detected on CXR and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on CXR, the contact is treated per the standard 9-month INH chemoprophylaxis protocol for LTBI. If TST induration diameter is <5 mm in non BCG-vaccinated contacts and <10 mm in BCG-vaccinated contacts, the contact has no CXR and no need for follow-up. We used published estimates of sensitivity and specificity of TST from a meta-analysis [14]. (4) TST followed by HRCT strategy: The contact undergoes TST testing. If TST induration diameter is ≥5 mm in non BCG-vaccinated contacts and ≥10 mm in BCG-vaccinated contacts, the contact undergoes HRCT. If TST is positive, active TB is detected on HRCT, and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on HRCT, the contact is treated per
the standard 9-month INH chemoprophylaxis protocol for LTBI. If TST induration diameter is <5 mm in non BCG-vaccinated contacts and <10 mm in BCG-vaccinated contacts, the contact has no HRCT and no need for follow-up. (5) CXR strategy: The contact undergoes CXR. If active TB is detected on CXR and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on CXR, the contact has no need for follow-up. (6) HRCT strategy: The contact undergoes HRCT. If active TB is detected on HRCT and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on HRCT, the contact has no need for follow-up.
2.3. Data sources, data, outcomes, and assumptions Using MEDLINE, we undertook a search of the literature published from 1980 to October 27, 2012. We assume the adherence rate (the proportion of patients who accept LTBI treatment) of the standard 9-month INH chemoprophylaxis protocol, the probability of INH-induced hepatitis, and the efficacy (preventing progression from LTBI to TB) of the standard 9-month chemoprophylaxis [9,10]. We also assume the probability of successful active TB treatment [15]. Age-specific all-cause mortality rates were obtained from Japanese life tables [16]. Age-specific TB-related mortality rates were obtained from the database from Japanese tuberculosis surveillance [17] and the excess cancer mortality risk by MDCT radiation exposure was considered [18] (Table 1). Data from meta-analyses were used for determining the sensitivities and specificities of QFT and TST (non BCG-vaccinated, BCG-vaccinated) [12,14]. The sensitivities and specificities of CXR and HRCT were obtained from studies conducted in numerous countries [1–4,13]. All costs were adjusted to 2012 Japanese yen, using the medical care component of the Japanese consumer price index [19], and were converted to US dollars, using the Organisation for Economic Co-operation and Development (OECD) purchasing power parity rate in 2009. Cost data were collected from various published sources [9,19,20]. The cost of QFT screening included the screening kits, one physician visit, and the labor cost for laboratory technicians. The cost of TST screening included the labor cost for the two physician visits and the TST reagents. The cost of CXR and HRCT included the material cost of CXR and HRCT, one physician visit, and the labor cost for radiological technicians. The costs of treating active TB, the standard 9-month INH chemoprevention, as well as the treatment of liver dysfunction was determined, based on published literature [9]. The costs of the smear and culture examinations of sputum were also considered when active TB was detected by CXR or HRCT. The main outcome measure of effectiveness was qualityadjusted life-years (QALYs) gained. QALY is the life expectancy which takes into account both the quantity and quality of life. Incremental cost, incremental effectiveness and incremental cost-effectiveness ratio (ICER) were calculated. Incremental cost is the increase or decrease in costs compared to the QFT followed by HRCT strategy. Incremental effectiveness is the increase or decrease in effectiveness compared to the QFT followed by HRCT strategy. ICER is the ratio of incremental cost per the incremental effectiveness. The lower ICERs correspond to better values. ICER of each screening arm was applied and compared. ‘Dominated’ means that the strategy costs more and is less effective than the QFT followed by HRCT strategy. All costs and clinical benefits were discounted at a fixed annual rate of 3%. A discount rate of 3% was applied by convention.
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
1355
Fig. 1. Simplified illustration of the decision trees for six strategies on tuberculosis contact screenings. A square node represents the decision node. A circle node represents a chance node. Branches from a chance node represent possible outcomes. An ‘M’ node represents a Markov node. HRCT = high resolution computed tomography; INH = standard 9-month INH chemoprophylaxis protocol for LTBI; QFT = QuantiFERON-TB Gold In-Tube; TST = tuberculin skin test; CXR = chest X-ray examination.
2.4. Sensitivity analyses Using one-way and two-way sensitivity analyses, the range of cost effectiveness was explored by comparing all strategies simultaneously to determine which strategy yielded the greatest benefits. Each model variable was assigned a distribution based on the values in the literature or assumptions. The type of distribution used for each variable, their median and 95% confidence interval (if available) are displayed in Table 1. Probabilistic sensitivity analyses recalculated expected values of each arm of the decision
analysis model 10,000 times. These results estimate the total impact of parameter uncertainty on the model. 3. Results 3.1. Base case In the base-case analysis, QFT followed by HRCT strategy yielded greatest benefits at the lowest cost ($US 6308.65; 27.56045 qualityadjusted life-years [QALYs]) [year 2012 values], and CXR alone
1356
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
Table 1 Baseline estimates for selected model variables. Baseline value Probability of having LTBI on contact investigation Probability of having TB on contact investigation Excess cancer mortality risk by MDCT radiation
0.281 0.014 0.0014
Mortality rate by active TB among TB patients Age 20–29 years Age 30–39 years Age 40–49 years Age 50–59 years Age 60–69 years Age 70–79 years Age 80 -100years
0.000595 0.000946 0.002349 0.008512 0.015800 0.050108 0.195661
Probability of developing TB from LTBI Age 20 –35 years Age 36–55 years Age 56-100years
0.0056 0.0042 0.0017
One-way sensitivity analysis range 0.242–0.324 0.011–0.018 0.001–0.019
References [27] [27] [18] [17]
[28]
Probability of successful TB treatment Probability of recurrence of active TB after treatment Adherence rate of standard 9-month INH chemoprophylaxis protocol when QFT is positive and CXR is negative Adherence rate of standard 9-month INH chemoprophylaxis protocol when TST is positive and CXR is negative Probability of INH-induced hepatitis by INH prophylaxis BCG vaccination rate Sensitivity of QFT for LTBI Specificity of QFT for LTBI Sensitivity of TST for LTBI Specificity of TST (BCG-vaccinated) for LTBI Specificity of TST (non BCG-vaccinated) for LTBI Sensitivity of CXR for active TB Specificity of CXR for active TB Sensitivity of HRCT for active TB Specificity of HRCT for active TB Cost ($US 2012 1$ = ¥83) QFT TST CXR HRCT Smear and culture of sputum examination Chemoprophylaxis by INH for 9 months Treatment of INH-induced hepatitis by INH chemoprophylaxis Treatment of TB for 6 months Average physician income per hour for physicians Average income per hour for radiology or laboratory technicians Utility Well LTBI LTBI taking chemoprophylaxis without complication LTBI taking chemoprophylaxis with complication Non-fatal active TB during treatment and before Dead
0.392 0.035 0.8
0.1–0.6 0.02–0.05 0.6–1.0
[15] [11] [9]
0.5
0.2–0.8
[9]
0.038 0.977 0.84 0.99 0.77 0.59 0.97 0.70 0.60 0.88 0.87
0.03–0.05 0–1 0.81–0.87 0.98–1.00 0.71–0.82 0.46–0.73 0.95–0.99 0.59–0.82 0.52–0.63 0.80–0.96 0.70–0.97
[10] [30] [12]
75.9 19.3 45.4 177.1 87.5 968.4 14900.2 18625.8 65.9 28.8
[14]
[13] [1–4]
38.0–151.8 9.6–38.6 22.7–90.8 88.6–354.2 43.8–175.0 484.2–1 936.9 7450.1–29800.3 9312.9–37251.5 32.9–131.8 14.4–57.6
1 1 0.99 0.85 0.80 0
[19]
[19] [9] [9] [9] [20]
[31] 0.98–1.00 0.80–0.90 0.75–0.85
LTBI = latent tuberculosis infection; TB = tuberculosis; MDCT = multi-director row computed tomography; HRCT = high resolution computed tomography; CXR = chest X-ray; INH = isonicotinyl hydrazide; TST = tuberculin skin test; QFT = QuantiFERON® -TB Gold In-Tube; BCG = bacillus Calmette-Guérin.
Table 2 Cost effectiveness of six strategies for TB contact screening. Strategy
Cost ($US 2012)
QFT+HRCT TST+HRCT QFT+CXR HRCT TST+CXR CXR
6308.65 7237.36 7418.60 8386.42 9589.37 13161.86
Incr Cost ($US 2012)
Effectiveness (QALY)
Incr Eff (QALY)
ICER ($US/QALY)
928.71 1109.95 2077.77 3280.72 6853.21
27.56045 27.38829 27.24563 27.14450 27.15881 25.86309
−0.17216 −0.31481 −0.41594 −0.40163 −1.69735
Dominated Dominated Dominated Dominated Dominated
Incr Cost = incremental cost; Incr Eff = incremental effectiveness; QALY = quality-adjusted life-year; ICER = incremental cost-effectiveness ratio; QFT = QuantiFERON® -TB Gold In-Tube; CXR = chest X-ray examination; HRCT = high resolution computed tomography; TST = tuberculin skin test.
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
1357
Fig. 2. One-way sensitivity analysis. TST followed by HRCT strategy was more costeffective than QFT followed by HRCT strategy at the willingness to pay level of $US 50,000/QALY gained when BCG vaccination rate was less than 0.12. NMB = net monetary benefit; HRCT = high resolution computed tomography; QFT = QuantiFERON-TB Gold In-Tube; TST = tuberculin skin test; BCG = bacillus Calmette-Guérin.
strategy was the least cost-effective ($US 13,161.86; 25.86309 QALYs) (Table 2). 3.2. One-way and two-way sensitivity analyses One-way and two-way sensitivity analyses were performed over the ranges for each variable. The cost-effectiveness was sensitive to BCG-vaccination rate. On one-way sensitivity analysis, TST followed by HRCT strategy yielded greater benefits at a lower cost than QFT followed by HRCT strategy at the willingness to pay level of $US 50,000/QALY gained when BCG-vaccination rate was less than 0.12 (Fig. 2). Net monetary benefit (NMB) was calculated: NMB = (Effectiveness × Willingness to pay [WTP]) − Cost. NMB is interpreted to costs with including the effectiveness expressed in terms of monetary units by taking the threshold of WTP into consideration. On two-way sensitivity analysis, TST followed by HRCT strategy was more cost-effective than QFT followed by HRCT strategy at the willingness to pay level of $US 50,000/QALY gained when adherence rate of chemoprevention in QFT strategy was 1.0 and BCG vaccination rate was less than 0.18 and when adherence rate of chemoprevention in QFT strategy was 0.6 and BCG vaccination rate was less than 0.05 (Fig. 3). 3.3. Probabilistic sensitivity analyses The cost-effectiveness acceptability curve of 20-year-olds contacts by Monte Carlo simulations for 10,000 trials demonstrated that QFT followed by HRCT was the most cost-effective, with a value of 100% at all willingness-to-pay levels. 4. Discussion In this study, we demonstrate that the QFT followed by HRCT strategy provides the greatest benefits at the lowest cost for TB contact investigations. The main reasons for dominance of QFT followed by HRCT are higher sensitivities and specificities of QFT and HRCT. Their higher sensitivities and specificities lead to decreased costs likely as a result of fewer false positives for both LTBI and active TB. Cost-effectiveness was sensitive to BCG-vaccination rate. TST followed by HRCT strategy was more cost-effective than QFT followed by HRCT strategy when BCG-vaccination rate was under 0.12. The lower BCG-vaccination rate is, the higher TST specificity is. TST specificity and low adherence rate of INH chemoprophylaxis of TST followed by HRCT strategy leads to higher effectiveness. We previously reported the cost-effectiveness of an IGRA compared to
Fig. 3. Two-way sensitivity analysis. TST followed by HRCT strategy was more costeffective than QFT followed by HRCT strategy at the willingness to pay level of $US 50,000/QALY gained when adherence rate of chemoprevention in QFT strategy was 1.0 and BCG vaccination rate was less than 0.18 and when adherence rate of chemoprevention in QFT strategy was 0.6 and BCG vaccination rate was less than 0.05. HRCT = high resolution computed tomography; QFT = QuantiFERON-TB Gold In-Tube; TST = tuberculin skin test; BCG = bacillus Calmette-Guérin.
TST and CXR for high-risk group screening and also showed that QFT alone yielded the greatest benefits at the lowest cost [21–25]. The most important reason for low accuracy of CXR in diagnosis of active TB is the non-specificity of CXR findings of TB. CXR has limited ability to characterize the imaging findings of TB. When active TB is misdiagnosed by CXR and chemoprophylaxis for LTBI using only INH is given to patients with active TB, there is an increased risk of developing drug resistant TB. HRCT has been demonstrated as a useful tool for assisting clinical decision about isolating patients pending sputum smear and/or nucleic acid confirmatory test results for active pulmonary TB [1–4]. HRCT has been shown to have higher sensitivity and specificity than CXR for detecting active pulmonary TB [1–4]. However, the increase of the radiation dose delivered in CT has recently been a problem. There is a trade-off between image quality and radiation dose on CT. With advances in computer hardware, various iterative reconstruction algorithms for CT images have been already used for dose reduction in clinical practice. Neroladaki et al. showed that model-based iterative reconstruction allows detection of pulmonary nodules with ultra-low-dose chest computed tomography with radiation exposure in the range of a posterior to anterior (PA) and lateral chest X-ray [5]. Katsura et al. demonstrated that diagnostically acceptable chest CT images acquired with nearly 80% less radiation can be obtained using Model-based iterative reconstruction [6]. Lee et al. suggested that the combined results of HRCT and the IGRA could help early diagnosis of active pulmonary TB and decision-making for early initiation of treatment in smear-negative patients [7]. CXR findings of TB bear a close resemblance to those of non-tuberculous mycobacterial pulmonary infection (NTM). The imaging findings of cavitary type of NTM are essentially the same with those of TB. Polverosi et al. showed that the HRCT pattern is very useful for differential diagnosis between TB and NTM, as well as using IGRAs [26]. The ability of HRCT to differentiate NTM and TB is still limited. To our knowledge, this study is the first cost-effectiveness analysis of HRCT for TB screening, comparing CXR using Markov models. Our study had several limitations. First, the parameters used in our model may be changeable in more precise investigations of TB dynamics, but there were little published data on this topic to
1358
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
base our estimates [27,28]. Second, we used estimates of sensitivities and specificities of HRCT and CXR for diagnosing active TB in our model. They were obtained from a limited number of papers [1–4,13]. Third, this study is limited in the model on immunocompetent individuals. The image findings of TB are also greatly influenced by the immune status of the patient. Further study is needed in immunosuppressive contacts such as human immunodeficiency virus infection/acquired immunodeficiency syndrome (HIV/AIDS) and pediatric patients. Fourth, the cost of transport for the contacts receiving HRCT and CXR was not considered in our model. Fifth, the installation fee of MDCT was not included into the calculation of HRCT cost. Finally, CT has still the risk of cancer induction with a 10–30 years latency time [18]. Excess cancer mortality risk by MDCT radiation was considered by the increase of all-cause mortality in our model. Radiologists should keep the efforts to reduce the radiation dose while maintaining tuberculosis diagnostic accuracy [29]. In conclusions, the QFT followed by HRCT strategy yielded the greatest benefits at the lowest cost for contact investigation. HRCT chest imaging, instead of CXR, is recommended as a cost effective addition to the evaluation and management of tuberculosis contacts in public health policy. Acknowledgments No sources of funding were used to conduct this study or prepare this manuscript. The author has no conflicts of interest that are directly relevant to the content of this study. References [1] Wang YH, Lin AS, Lai YF, Chao TY, Liu JW, Ko SF. The high value of high-resolution computed tomography in predicting the activity of pulmonary tuberculosis. International Journal of Tuberculosis and Lung Disease 2003;7(6):563–8. [2] Tozkoparan E, Deniz O, Ciftici F, et al. The roles of HRCT and clinical parameters in assessing activity of suspected smear negative pulmonary tuberculosis. Archives of Medical Research 2005;36(2):166–70. [3] Yeh JJ, Yu JK, Teng WB, et al. High-resolution CT for identify patients with smear-positive, active pulmonary tuberculosis. European Journal of Radiology 2012;81(1):195–201. [4] Yeh JJ, Chen SC, Teng WB, et al. Identifying the most infectious lesions in pulmonary tuberculosis by high-resolution multi-detector computed tomography. European Journal of Radiology 2010;20(9):2135–45. [5] Neroladaki A, Botsikas D, Boudabbous S, Becker CD, Montet X. Computed tomography of the chest with model-based iterative reconstruction using a radiation exposure similar to chest X-ray examination: preliminary observations. European Radiology 2013;23(2):360–6. [6] Katsura M, Matsuda I, Akahane M, et al. Model-based iterative reconstruction technique for radiation dose reduction in chest CT: comparison with the adaptive statistical iterative reconstruction technique. European Radiology 2012;22(8):1613–23. [7] Lee HM, Shin JW, Kim JY, et al. HRCT and whole-blood interferon-gamma assay for the rapid diagnosis of smear-negative pulmonary tuberculosis. Respiration 2010;79(6):454–60. [8] Eisenberg JM. Clinical economics. A guide to the economic analysis of clinical practices. JAMA 1989;262(20):2879–86. [9] Yoshiyama T. Cost effectiveness analysis of isoniazid preventive therapy to the contacts of tuberculosis patients under Japanese settings. Kekkaku 2000;75(11):629–41 [in Japanese].
[10] Fountain FF, Tolley E, Chrisman CR, et al. Isoniazid hepatotoxicity associated with treatment of latent tuberculosis infection: a 7-year evaluation from a public health tuberculosis clinic. Chest 2005;128(1):116–23. [11] Tuberculosis Research Committee. Relapse rate of tuberculosis treated with standard regimen of chemotherapy. Kekkaku 2009;84:617–25 [in Japanese]. [12] Diel R, Loddenkemper R, Nienhaus A. Evidence-based comparison of commercial interferon-gamma release assays for detecting active TB: a metaanalysis. Chest 2010;137:952–68. [13] Tattevin P, Casalino E, Fleury L, et al. The validity of medical history, classic symptoms, and chest radiographs in predicting pulmonary tuberculosis: derivation of a pulmonary tuberculosis prediction model. Chest 1999;115(5):1248–53. [14] Pai M, Zwerling A, Menzies D. Systematic review: T-cell-based assays for the diagnosis of latent tuberculosis infection: an update. Annals of Internal Medicine 2008;149(3):177–84. [15] Chee CB, KhinMar KW, Gan SH, et al. Tuberculosis treatment effect on T-cell interferon-gamma responses to Mycobacterium tuberculosis-specific antigens. European Respiratory Journal 2010;36(2):355–61. [16] Statistics Bureau, Ministry of Internal Affairs and Communications. Life tables, 2009. Tokyo: Statistics Bureau, 2010 [online]. Available from URL: http://www. mhlw.go.jp/english/database/db-hw/vs02.html [accessed 27.10.12]. [17] Research Institute of Tuberculosis. Tuberculosis surveillance 2009 [in Japanese]. Tokyo: Research Institute of Tuberculosis, 2009 [online]. Available from URL: http://www.jata.or.jp/rit/ekigaku/ [accessed 27.10.12]. [18] Brenner DJ, Elliston CD. Estimated radiation risks potentially associated with full-body CT screening. Radiology 2004;232(3):735–8. [19] Igakutsushin-sya. Medical insurance reimbursement table in Japan [in Japanese]. Tokyo: Igakutsushin-sya, 2010. [20] Ministry of Health, Labor and Welfare. Basic survey on wage structure [online]. Available from URL: http://www.mhlw.go.jp/english/database/ db-l/index.html [accessed 27.10.12]. [21] Kowada A. Cost-effectiveness of interferon-␥ release assay versus chest X-ray for tuberculosis screening of employees. American Journal of Infection Control 2011;39(10):e67–72. [22] Kowada A. Cost effectiveness of interferon-gamma release assay for school-based tuberculosis screening. Molecular Diagnosis and Therapy 2012;16(3):181–90. [23] Kowada A, Deshpande GA, Takahashi O, Shimbo T, Fukui T. Cost-effectiveness analysis of interferon-␥ release assays versus chest X-ray for annual tuberculosis screening of healthcare workers. Journal of Hospital Infection 2011;78(2):152–4. [24] Kowada A, Deshpande GA, Takahashi O, et al. Cost effectiveness of interferon-gamma release assay versus chest X-ray for tuberculosis screening of BCG-vaccinated elderly populations. Molecular Diagnosis and Therapy 2010;14(4):229–36. [25] Kowada A. Cost effectiveness of interferon-gamma release assay for tuberculosis screening of rheumatoid arthritis patients prior to initiation of tumor necrosis factor-␣ antagonist therapy. Molecular Diagnosis and Therapy 2010;14(6):367–73. [26] Polverosi R, Guarise A, Balestro E, et al. High-resolution CT of nontuberculous mycobacteria pulmonary infection in immunocompetent, non-HIV-positive patients. Radiologia Medica 2010;115(2):191–204. [27] Fox GJ, Barry SE, Britton WJ, Marks GB. Contact investigation for tuberculosis: a systematic review and meta-analysis. European Respiratory Journal 2013;41(1):140–56. [28] Horsburgh Jr CR. Priorities for the treatment of latent tuberculosis infection in the United States. New England Journal of Medicine 2004;350(20): 2060–7. [29] Mayo JR, Aldrich J, Muller NL. Fleischner Society. Radiation exposure at chest CT: a statement of the Fleischner Society. Radiology 2003;228(1):15–21. Review. [30] Takayama N, Sakiyama H, Okabe N, Umemoto S. Cumulative national immunization rate of BCG vaccine in Japan. Journal of Japan Medical Association 2009;137:2132–6 [in Japanese]. [31] Marra CA, Marra F, Colley L, et al. Health-related quality of life trajectories among adults with tuberculosis: differences between latent and active infection. Chest 2008;133:396–403.
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Cost effectiveness of high resolution computed tomography with interferon-gamma release assay for tuberculosis contact investigation Akiko Kowada ∗ Kojiya Haneda Healthcare Service, Ota City Public Health Office, Tokyo, Japan
a r t i c l e
i n f o
Article history: Received 30 October 2012 Received in revised form 2 February 2013 Accepted 9 February 2013 Keywords: High resolution computed tomography Interferon-gamma release assay Cost-effectiveness Tuberculosis contacts Chest X-ray
a b s t r a c t Background: Tuberculosis contact investigation is one of the important public health strategies to control tuberculosis worldwide. Recently, high resolution computed tomography (HRCT) has been reported as a more accurate radiological method with higher sensitivity and specificity than chest X-ray (CXR) to detect active tuberculosis. In this study, we assessed the cost effectiveness of HRCT compared to CXR in combination with QuantiFERON® -TB Gold In-Tube (QFT) or the tuberculin skin test (TST) for tuberculosis contact investigation. Methods: We constructed Markov models using a societal perspective on the lifetime horizon. The target population was a hypothetical cohort of immunocompetent 20-year-old contacts with smear-positive tuberculosis patients in developed countries. Six strategies; QFT followed by CXR, QFT followed by HRCT, TST followed by CXR, TST followed by HRCT, CXR alone and HRCT alone were modeled. All costs and clinical benefits were discounted at a fixed annual rate of 3%. Results: In the base-case analysis, QFT followed by HRCT strategy yielded the greatest benefit at the lowest cost ($US 6308.65; 27.56045 quality-adjusted life-years [QALYs])[year 2012 values]. Cost-effectiveness was sensitive to BCG vaccination rate. Conclusions: The QFT followed by HRCT strategy yielded the greatest benefits at the lowest cost. HRCT chest imaging, instead of CXR, is recommended as a cost effective addition to the evaluation and management of tuberculosis contacts in public health policy. © 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Tuberculosis (TB) is a widespread infectious and serious disease for global public health. To control TB, TB contact investigations play a very important role in public health policies not only by diagnosing latent tuberculosis infection (LTBI) and allowing chemoprophylaxis but also by earlier detection of active TB, often leading to treatment before appearance of symptoms of TB. The main approach for detecting active pulmonary TB is currently radiological examination by chest X-ray (CXR). However, low sensitivity and specificity of CXR are well known limitations. Recently, chest computed tomography has been reported as a more accurate and effective radiological method with higher sensitivity and specificity [1–4]. Multi-detector row computed tomography (MDCT) has greatly increased the clinical indications for CT and promotes to minimize excessive CT radiation exposure. High resolution CT
∗ Tel.: +81 3 3743 8315; fax: +81 3 3742 3116. E-mail address: [email protected] 0720-048X/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejrad.2013.02.017
(HRCT) images can be created from raw-data of MDCT without additional radiation [5–7]. Mycobacterium tuberculosis-specific interferon-gamma release assays (IGRAs) – QuantiFERON® TB Gold In-Tube (QFT) [Cellestis Limited, Chadstone, VIC, Australia] and T-SPOT® .TB (Oxford Immunotec, Oxford, UK) are now available and provide more accurate and sensitive diagnosis of M. tuberculosis infection with higher specificity than that of the tuberculin skin test (TST). They are not affected by bacillus Calmette-Guérin (BCG) vaccination and do not suffer from the booster phenomenon seen with repeated TSTs. However, these tests do not discriminate between active TB and LTBI and thus those testing positive are usually screened for active pulmonary TB using CXR. There are recent reports demonstrating the utility of using HRCT, rather than CXR, in combination with IGRAs for tuberculosis screening [7]. However, although more accurate, the purchase costs of HRCT and IGRAs are higher than CXR and TST and thus their overall cost effectiveness as mass TB screening tools warrants evaluation. In this study, we assessed the cost effectiveness of HRCT versus CXR in combination with QFT or TST strategies for TB contact investigation to demonstrate the optimal screening method for contacts.
1354
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
2. Methods 2.1. Target population Immunocompetent 20-year-olds contacts were chosen as a hypothetical cohort on lifetime horizon in developed countries. 2.2. Markov models The following four clinical states were included in our model to represent the possible clinical states in the target population: (i) Well (no LTBI and no TB); (ii) LTBI; (iii) TB; (iv) Death. Decision-analytical calculations were performed using Tree Age Pro Healthcare Module 2009 (Tree Age Software Inc., Williamstown, MA, USA). Each cycle length was 1 year. As this was a modeling study with all inputs and parameters derived from published literatures, ethics approval was not required. Markov models were developed for six strategies: QFT followed by CXR, QFT followed by HRCT, TST followed by CXR, TST followed by HRCT, CXR alone and HRCT alone (Fig. 1). The contacts were stratified by BCG-vaccination status for TST strategies. Markov models are state transition models to calculate expected values, costs and utilities [8]. (1) QFT followed by CXR strategy: The contact undergoes QFT testing. If QFT is positive, active TB is detected on CXR, and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If QFT is positive and active TB is not detected on CXR, the contact is treated per the standard 9-month isonicotinyl hydrazide (INH) chemoprophylaxis regimen for LTBI. If QFT is negative, the contact has no CXR and no need for follow-up. We considered the adherence and complication rates of chemoprevention [9,10]. We also estimated the recurrence of TB [11]. We used published estimates of sensitivity and specificity of QFT from a meta-analysis of developed country studies [12]. We used published estimates of sensitivity and specificity of CXR [13]. (2) QFT followed by HRCT strategy: The contact undergoes QFT testing. If QFT is positive, active TB is detected on HRCT, and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on HRCT, the contact is treated per the standard 9-month INH chemoprophylaxis for LTBI. If QFT is negative, the contact has no HRCT and no need for follow-up. We used published estimates of sensitivity and specificity of HRCT [1–4]. (3) TST followed by CXR strategy: The contact undergoes TST testing. If TST induration diameter is ≥5 mm in non BCG-vaccinated contacts and ≥10 mm in BCG-vaccinated contacts, the contact undergoes CXR. If active TB is detected on CXR and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on CXR, the contact is treated per the standard 9-month INH chemoprophylaxis protocol for LTBI. If TST induration diameter is <5 mm in non BCG-vaccinated contacts and <10 mm in BCG-vaccinated contacts, the contact has no CXR and no need for follow-up. We used published estimates of sensitivity and specificity of TST from a meta-analysis [14]. (4) TST followed by HRCT strategy: The contact undergoes TST testing. If TST induration diameter is ≥5 mm in non BCG-vaccinated contacts and ≥10 mm in BCG-vaccinated contacts, the contact undergoes HRCT. If TST is positive, active TB is detected on HRCT, and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on HRCT, the contact is treated per
the standard 9-month INH chemoprophylaxis protocol for LTBI. If TST induration diameter is <5 mm in non BCG-vaccinated contacts and <10 mm in BCG-vaccinated contacts, the contact has no HRCT and no need for follow-up. (5) CXR strategy: The contact undergoes CXR. If active TB is detected on CXR and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on CXR, the contact has no need for follow-up. (6) HRCT strategy: The contact undergoes HRCT. If active TB is detected on HRCT and the sputum smears and/or cultures are positive, the contact is treated per the standard 6-month protocol for active TB. If active TB is not detected on HRCT, the contact has no need for follow-up.
2.3. Data sources, data, outcomes, and assumptions Using MEDLINE, we undertook a search of the literature published from 1980 to October 27, 2012. We assume the adherence rate (the proportion of patients who accept LTBI treatment) of the standard 9-month INH chemoprophylaxis protocol, the probability of INH-induced hepatitis, and the efficacy (preventing progression from LTBI to TB) of the standard 9-month chemoprophylaxis [9,10]. We also assume the probability of successful active TB treatment [15]. Age-specific all-cause mortality rates were obtained from Japanese life tables [16]. Age-specific TB-related mortality rates were obtained from the database from Japanese tuberculosis surveillance [17] and the excess cancer mortality risk by MDCT radiation exposure was considered [18] (Table 1). Data from meta-analyses were used for determining the sensitivities and specificities of QFT and TST (non BCG-vaccinated, BCG-vaccinated) [12,14]. The sensitivities and specificities of CXR and HRCT were obtained from studies conducted in numerous countries [1–4,13]. All costs were adjusted to 2012 Japanese yen, using the medical care component of the Japanese consumer price index [19], and were converted to US dollars, using the Organisation for Economic Co-operation and Development (OECD) purchasing power parity rate in 2009. Cost data were collected from various published sources [9,19,20]. The cost of QFT screening included the screening kits, one physician visit, and the labor cost for laboratory technicians. The cost of TST screening included the labor cost for the two physician visits and the TST reagents. The cost of CXR and HRCT included the material cost of CXR and HRCT, one physician visit, and the labor cost for radiological technicians. The costs of treating active TB, the standard 9-month INH chemoprevention, as well as the treatment of liver dysfunction was determined, based on published literature [9]. The costs of the smear and culture examinations of sputum were also considered when active TB was detected by CXR or HRCT. The main outcome measure of effectiveness was qualityadjusted life-years (QALYs) gained. QALY is the life expectancy which takes into account both the quantity and quality of life. Incremental cost, incremental effectiveness and incremental cost-effectiveness ratio (ICER) were calculated. Incremental cost is the increase or decrease in costs compared to the QFT followed by HRCT strategy. Incremental effectiveness is the increase or decrease in effectiveness compared to the QFT followed by HRCT strategy. ICER is the ratio of incremental cost per the incremental effectiveness. The lower ICERs correspond to better values. ICER of each screening arm was applied and compared. ‘Dominated’ means that the strategy costs more and is less effective than the QFT followed by HRCT strategy. All costs and clinical benefits were discounted at a fixed annual rate of 3%. A discount rate of 3% was applied by convention.
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
1355
Fig. 1. Simplified illustration of the decision trees for six strategies on tuberculosis contact screenings. A square node represents the decision node. A circle node represents a chance node. Branches from a chance node represent possible outcomes. An ‘M’ node represents a Markov node. HRCT = high resolution computed tomography; INH = standard 9-month INH chemoprophylaxis protocol for LTBI; QFT = QuantiFERON-TB Gold In-Tube; TST = tuberculin skin test; CXR = chest X-ray examination.
2.4. Sensitivity analyses Using one-way and two-way sensitivity analyses, the range of cost effectiveness was explored by comparing all strategies simultaneously to determine which strategy yielded the greatest benefits. Each model variable was assigned a distribution based on the values in the literature or assumptions. The type of distribution used for each variable, their median and 95% confidence interval (if available) are displayed in Table 1. Probabilistic sensitivity analyses recalculated expected values of each arm of the decision
analysis model 10,000 times. These results estimate the total impact of parameter uncertainty on the model. 3. Results 3.1. Base case In the base-case analysis, QFT followed by HRCT strategy yielded greatest benefits at the lowest cost ($US 6308.65; 27.56045 qualityadjusted life-years [QALYs]) [year 2012 values], and CXR alone
1356
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
Table 1 Baseline estimates for selected model variables. Baseline value Probability of having LTBI on contact investigation Probability of having TB on contact investigation Excess cancer mortality risk by MDCT radiation
0.281 0.014 0.0014
Mortality rate by active TB among TB patients Age 20–29 years Age 30–39 years Age 40–49 years Age 50–59 years Age 60–69 years Age 70–79 years Age 80 -100years
0.000595 0.000946 0.002349 0.008512 0.015800 0.050108 0.195661
Probability of developing TB from LTBI Age 20 –35 years Age 36–55 years Age 56-100years
0.0056 0.0042 0.0017
One-way sensitivity analysis range 0.242–0.324 0.011–0.018 0.001–0.019
References [27] [27] [18] [17]
[28]
Probability of successful TB treatment Probability of recurrence of active TB after treatment Adherence rate of standard 9-month INH chemoprophylaxis protocol when QFT is positive and CXR is negative Adherence rate of standard 9-month INH chemoprophylaxis protocol when TST is positive and CXR is negative Probability of INH-induced hepatitis by INH prophylaxis BCG vaccination rate Sensitivity of QFT for LTBI Specificity of QFT for LTBI Sensitivity of TST for LTBI Specificity of TST (BCG-vaccinated) for LTBI Specificity of TST (non BCG-vaccinated) for LTBI Sensitivity of CXR for active TB Specificity of CXR for active TB Sensitivity of HRCT for active TB Specificity of HRCT for active TB Cost ($US 2012 1$ = ¥83) QFT TST CXR HRCT Smear and culture of sputum examination Chemoprophylaxis by INH for 9 months Treatment of INH-induced hepatitis by INH chemoprophylaxis Treatment of TB for 6 months Average physician income per hour for physicians Average income per hour for radiology or laboratory technicians Utility Well LTBI LTBI taking chemoprophylaxis without complication LTBI taking chemoprophylaxis with complication Non-fatal active TB during treatment and before Dead
0.392 0.035 0.8
0.1–0.6 0.02–0.05 0.6–1.0
[15] [11] [9]
0.5
0.2–0.8
[9]
0.038 0.977 0.84 0.99 0.77 0.59 0.97 0.70 0.60 0.88 0.87
0.03–0.05 0–1 0.81–0.87 0.98–1.00 0.71–0.82 0.46–0.73 0.95–0.99 0.59–0.82 0.52–0.63 0.80–0.96 0.70–0.97
[10] [30] [12]
75.9 19.3 45.4 177.1 87.5 968.4 14900.2 18625.8 65.9 28.8
[14]
[13] [1–4]
38.0–151.8 9.6–38.6 22.7–90.8 88.6–354.2 43.8–175.0 484.2–1 936.9 7450.1–29800.3 9312.9–37251.5 32.9–131.8 14.4–57.6
1 1 0.99 0.85 0.80 0
[19]
[19] [9] [9] [9] [20]
[31] 0.98–1.00 0.80–0.90 0.75–0.85
LTBI = latent tuberculosis infection; TB = tuberculosis; MDCT = multi-director row computed tomography; HRCT = high resolution computed tomography; CXR = chest X-ray; INH = isonicotinyl hydrazide; TST = tuberculin skin test; QFT = QuantiFERON® -TB Gold In-Tube; BCG = bacillus Calmette-Guérin.
Table 2 Cost effectiveness of six strategies for TB contact screening. Strategy
Cost ($US 2012)
QFT+HRCT TST+HRCT QFT+CXR HRCT TST+CXR CXR
6308.65 7237.36 7418.60 8386.42 9589.37 13161.86
Incr Cost ($US 2012)
Effectiveness (QALY)
Incr Eff (QALY)
ICER ($US/QALY)
928.71 1109.95 2077.77 3280.72 6853.21
27.56045 27.38829 27.24563 27.14450 27.15881 25.86309
−0.17216 −0.31481 −0.41594 −0.40163 −1.69735
Dominated Dominated Dominated Dominated Dominated
Incr Cost = incremental cost; Incr Eff = incremental effectiveness; QALY = quality-adjusted life-year; ICER = incremental cost-effectiveness ratio; QFT = QuantiFERON® -TB Gold In-Tube; CXR = chest X-ray examination; HRCT = high resolution computed tomography; TST = tuberculin skin test.
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
1357
Fig. 2. One-way sensitivity analysis. TST followed by HRCT strategy was more costeffective than QFT followed by HRCT strategy at the willingness to pay level of $US 50,000/QALY gained when BCG vaccination rate was less than 0.12. NMB = net monetary benefit; HRCT = high resolution computed tomography; QFT = QuantiFERON-TB Gold In-Tube; TST = tuberculin skin test; BCG = bacillus Calmette-Guérin.
strategy was the least cost-effective ($US 13,161.86; 25.86309 QALYs) (Table 2). 3.2. One-way and two-way sensitivity analyses One-way and two-way sensitivity analyses were performed over the ranges for each variable. The cost-effectiveness was sensitive to BCG-vaccination rate. On one-way sensitivity analysis, TST followed by HRCT strategy yielded greater benefits at a lower cost than QFT followed by HRCT strategy at the willingness to pay level of $US 50,000/QALY gained when BCG-vaccination rate was less than 0.12 (Fig. 2). Net monetary benefit (NMB) was calculated: NMB = (Effectiveness × Willingness to pay [WTP]) − Cost. NMB is interpreted to costs with including the effectiveness expressed in terms of monetary units by taking the threshold of WTP into consideration. On two-way sensitivity analysis, TST followed by HRCT strategy was more cost-effective than QFT followed by HRCT strategy at the willingness to pay level of $US 50,000/QALY gained when adherence rate of chemoprevention in QFT strategy was 1.0 and BCG vaccination rate was less than 0.18 and when adherence rate of chemoprevention in QFT strategy was 0.6 and BCG vaccination rate was less than 0.05 (Fig. 3). 3.3. Probabilistic sensitivity analyses The cost-effectiveness acceptability curve of 20-year-olds contacts by Monte Carlo simulations for 10,000 trials demonstrated that QFT followed by HRCT was the most cost-effective, with a value of 100% at all willingness-to-pay levels. 4. Discussion In this study, we demonstrate that the QFT followed by HRCT strategy provides the greatest benefits at the lowest cost for TB contact investigations. The main reasons for dominance of QFT followed by HRCT are higher sensitivities and specificities of QFT and HRCT. Their higher sensitivities and specificities lead to decreased costs likely as a result of fewer false positives for both LTBI and active TB. Cost-effectiveness was sensitive to BCG-vaccination rate. TST followed by HRCT strategy was more cost-effective than QFT followed by HRCT strategy when BCG-vaccination rate was under 0.12. The lower BCG-vaccination rate is, the higher TST specificity is. TST specificity and low adherence rate of INH chemoprophylaxis of TST followed by HRCT strategy leads to higher effectiveness. We previously reported the cost-effectiveness of an IGRA compared to
Fig. 3. Two-way sensitivity analysis. TST followed by HRCT strategy was more costeffective than QFT followed by HRCT strategy at the willingness to pay level of $US 50,000/QALY gained when adherence rate of chemoprevention in QFT strategy was 1.0 and BCG vaccination rate was less than 0.18 and when adherence rate of chemoprevention in QFT strategy was 0.6 and BCG vaccination rate was less than 0.05. HRCT = high resolution computed tomography; QFT = QuantiFERON-TB Gold In-Tube; TST = tuberculin skin test; BCG = bacillus Calmette-Guérin.
TST and CXR for high-risk group screening and also showed that QFT alone yielded the greatest benefits at the lowest cost [21–25]. The most important reason for low accuracy of CXR in diagnosis of active TB is the non-specificity of CXR findings of TB. CXR has limited ability to characterize the imaging findings of TB. When active TB is misdiagnosed by CXR and chemoprophylaxis for LTBI using only INH is given to patients with active TB, there is an increased risk of developing drug resistant TB. HRCT has been demonstrated as a useful tool for assisting clinical decision about isolating patients pending sputum smear and/or nucleic acid confirmatory test results for active pulmonary TB [1–4]. HRCT has been shown to have higher sensitivity and specificity than CXR for detecting active pulmonary TB [1–4]. However, the increase of the radiation dose delivered in CT has recently been a problem. There is a trade-off between image quality and radiation dose on CT. With advances in computer hardware, various iterative reconstruction algorithms for CT images have been already used for dose reduction in clinical practice. Neroladaki et al. showed that model-based iterative reconstruction allows detection of pulmonary nodules with ultra-low-dose chest computed tomography with radiation exposure in the range of a posterior to anterior (PA) and lateral chest X-ray [5]. Katsura et al. demonstrated that diagnostically acceptable chest CT images acquired with nearly 80% less radiation can be obtained using Model-based iterative reconstruction [6]. Lee et al. suggested that the combined results of HRCT and the IGRA could help early diagnosis of active pulmonary TB and decision-making for early initiation of treatment in smear-negative patients [7]. CXR findings of TB bear a close resemblance to those of non-tuberculous mycobacterial pulmonary infection (NTM). The imaging findings of cavitary type of NTM are essentially the same with those of TB. Polverosi et al. showed that the HRCT pattern is very useful for differential diagnosis between TB and NTM, as well as using IGRAs [26]. The ability of HRCT to differentiate NTM and TB is still limited. To our knowledge, this study is the first cost-effectiveness analysis of HRCT for TB screening, comparing CXR using Markov models. Our study had several limitations. First, the parameters used in our model may be changeable in more precise investigations of TB dynamics, but there were little published data on this topic to
1358
A. Kowada / European Journal of Radiology 82 (2013) 1353–1358
base our estimates [27,28]. Second, we used estimates of sensitivities and specificities of HRCT and CXR for diagnosing active TB in our model. They were obtained from a limited number of papers [1–4,13]. Third, this study is limited in the model on immunocompetent individuals. The image findings of TB are also greatly influenced by the immune status of the patient. Further study is needed in immunosuppressive contacts such as human immunodeficiency virus infection/acquired immunodeficiency syndrome (HIV/AIDS) and pediatric patients. Fourth, the cost of transport for the contacts receiving HRCT and CXR was not considered in our model. Fifth, the installation fee of MDCT was not included into the calculation of HRCT cost. Finally, CT has still the risk of cancer induction with a 10–30 years latency time [18]. Excess cancer mortality risk by MDCT radiation was considered by the increase of all-cause mortality in our model. Radiologists should keep the efforts to reduce the radiation dose while maintaining tuberculosis diagnostic accuracy [29]. In conclusions, the QFT followed by HRCT strategy yielded the greatest benefits at the lowest cost for contact investigation. HRCT chest imaging, instead of CXR, is recommended as a cost effective addition to the evaluation and management of tuberculosis contacts in public health policy. Acknowledgments No sources of funding were used to conduct this study or prepare this manuscript. The author has no conflicts of interest that are directly relevant to the content of this study. References [1] Wang YH, Lin AS, Lai YF, Chao TY, Liu JW, Ko SF. The high value of high-resolution computed tomography in predicting the activity of pulmonary tuberculosis. International Journal of Tuberculosis and Lung Disease 2003;7(6):563–8. [2] Tozkoparan E, Deniz O, Ciftici F, et al. The roles of HRCT and clinical parameters in assessing activity of suspected smear negative pulmonary tuberculosis. Archives of Medical Research 2005;36(2):166–70. [3] Yeh JJ, Yu JK, Teng WB, et al. High-resolution CT for identify patients with smear-positive, active pulmonary tuberculosis. European Journal of Radiology 2012;81(1):195–201. [4] Yeh JJ, Chen SC, Teng WB, et al. Identifying the most infectious lesions in pulmonary tuberculosis by high-resolution multi-detector computed tomography. European Journal of Radiology 2010;20(9):2135–45. [5] Neroladaki A, Botsikas D, Boudabbous S, Becker CD, Montet X. Computed tomography of the chest with model-based iterative reconstruction using a radiation exposure similar to chest X-ray examination: preliminary observations. European Radiology 2013;23(2):360–6. [6] Katsura M, Matsuda I, Akahane M, et al. Model-based iterative reconstruction technique for radiation dose reduction in chest CT: comparison with the adaptive statistical iterative reconstruction technique. European Radiology 2012;22(8):1613–23. [7] Lee HM, Shin JW, Kim JY, et al. HRCT and whole-blood interferon-gamma assay for the rapid diagnosis of smear-negative pulmonary tuberculosis. Respiration 2010;79(6):454–60. [8] Eisenberg JM. Clinical economics. A guide to the economic analysis of clinical practices. JAMA 1989;262(20):2879–86. [9] Yoshiyama T. Cost effectiveness analysis of isoniazid preventive therapy to the contacts of tuberculosis patients under Japanese settings. Kekkaku 2000;75(11):629–41 [in Japanese].
[10] Fountain FF, Tolley E, Chrisman CR, et al. Isoniazid hepatotoxicity associated with treatment of latent tuberculosis infection: a 7-year evaluation from a public health tuberculosis clinic. Chest 2005;128(1):116–23. [11] Tuberculosis Research Committee. Relapse rate of tuberculosis treated with standard regimen of chemotherapy. Kekkaku 2009;84:617–25 [in Japanese]. [12] Diel R, Loddenkemper R, Nienhaus A. Evidence-based comparison of commercial interferon-gamma release assays for detecting active TB: a metaanalysis. Chest 2010;137:952–68. [13] Tattevin P, Casalino E, Fleury L, et al. The validity of medical history, classic symptoms, and chest radiographs in predicting pulmonary tuberculosis: derivation of a pulmonary tuberculosis prediction model. Chest 1999;115(5):1248–53. [14] Pai M, Zwerling A, Menzies D. Systematic review: T-cell-based assays for the diagnosis of latent tuberculosis infection: an update. Annals of Internal Medicine 2008;149(3):177–84. [15] Chee CB, KhinMar KW, Gan SH, et al. Tuberculosis treatment effect on T-cell interferon-gamma responses to Mycobacterium tuberculosis-specific antigens. European Respiratory Journal 2010;36(2):355–61. [16] Statistics Bureau, Ministry of Internal Affairs and Communications. Life tables, 2009. Tokyo: Statistics Bureau, 2010 [online]. Available from URL: http://www. mhlw.go.jp/english/database/db-hw/vs02.html [accessed 27.10.12]. [17] Research Institute of Tuberculosis. Tuberculosis surveillance 2009 [in Japanese]. Tokyo: Research Institute of Tuberculosis, 2009 [online]. Available from URL: http://www.jata.or.jp/rit/ekigaku/ [accessed 27.10.12]. [18] Brenner DJ, Elliston CD. Estimated radiation risks potentially associated with full-body CT screening. Radiology 2004;232(3):735–8. [19] Igakutsushin-sya. Medical insurance reimbursement table in Japan [in Japanese]. Tokyo: Igakutsushin-sya, 2010. [20] Ministry of Health, Labor and Welfare. Basic survey on wage structure [online]. Available from URL: http://www.mhlw.go.jp/english/database/ db-l/index.html [accessed 27.10.12]. [21] Kowada A. Cost-effectiveness of interferon-␥ release assay versus chest X-ray for tuberculosis screening of employees. American Journal of Infection Control 2011;39(10):e67–72. [22] Kowada A. Cost effectiveness of interferon-gamma release assay for school-based tuberculosis screening. Molecular Diagnosis and Therapy 2012;16(3):181–90. [23] Kowada A, Deshpande GA, Takahashi O, Shimbo T, Fukui T. Cost-effectiveness analysis of interferon-␥ release assays versus chest X-ray for annual tuberculosis screening of healthcare workers. Journal of Hospital Infection 2011;78(2):152–4. [24] Kowada A, Deshpande GA, Takahashi O, et al. Cost effectiveness of interferon-gamma release assay versus chest X-ray for tuberculosis screening of BCG-vaccinated elderly populations. Molecular Diagnosis and Therapy 2010;14(4):229–36. [25] Kowada A. Cost effectiveness of interferon-gamma release assay for tuberculosis screening of rheumatoid arthritis patients prior to initiation of tumor necrosis factor-␣ antagonist therapy. Molecular Diagnosis and Therapy 2010;14(6):367–73. [26] Polverosi R, Guarise A, Balestro E, et al. High-resolution CT of nontuberculous mycobacteria pulmonary infection in immunocompetent, non-HIV-positive patients. Radiologia Medica 2010;115(2):191–204. [27] Fox GJ, Barry SE, Britton WJ, Marks GB. Contact investigation for tuberculosis: a systematic review and meta-analysis. European Respiratory Journal 2013;41(1):140–56. [28] Horsburgh Jr CR. Priorities for the treatment of latent tuberculosis infection in the United States. New England Journal of Medicine 2004;350(20): 2060–7. [29] Mayo JR, Aldrich J, Muller NL. Fleischner Society. Radiation exposure at chest CT: a statement of the Fleischner Society. Radiology 2003;228(1):15–21. Review. [30] Takayama N, Sakiyama H, Okabe N, Umemoto S. Cumulative national immunization rate of BCG vaccine in Japan. Journal of Japan Medical Association 2009;137:2132–6 [in Japanese]. [31] Marra CA, Marra F, Colley L, et al. Health-related quality of life trajectories among adults with tuberculosis: differences between latent and active infection. Chest 2008;133:396–403.