The Cost-Utility Analysis of PET-Scan in Diagnosis and Treatment of Non-Small Cell Lung Carcinoma in Iran

Background PET scan is a non-invasive, complex and expensive medical imaging technology that is normally used for the diagnosis and treatment of various diseases including lung cancer. Objectives The purpose of this study is to assess the cost effectiveness of this technology in the diagnosis and treatment of non- small cell lung carcinoma (NSCLC) in Iran. Materials and Methods The main electronic databases including The Cochrane Library and Medline were searched to identify available evidence about the performance and effectiveness of technology. A standard decision tree model with seven strategies was used to perform the economic evaluation. Retrieved studies and expert opinion were used to estimate the cost of each treatment strategy in Iran. The costs were divided into three categories including capital costs (depreciation costs of buildings and equipment), staff costs and other expenses (including cost of consumables, running and maintenance costs). The costs were estimated in both IR-Rials and US-Dollars with an exchange rate of 10.000 IR Rials per one US Dollar according to the exchange rate in 2008. Results The total annual running cost of a PET scan was about 8850 to 13000 million Rials, (0.9 to 1.3 million US$). The average cost of performing a PET scan varied between 3 and 4.5 million Rials (300 to 450US$). The strategies 3 (mediastinoscopy alone) and 7 (mediastinoscopy after PET scan) were more cost-effective than other strategies, especially when the result of the CT-scan performed before PET scan was negative. Conclusion The technical performance of PET scan is significantly higher than similar technologies for staging and treatment of NSCLC. In addition, it might slightly improve the treatment process and lead to a small level of increase in the quality adjusted life year (QALY) gained by these patients making it cost-effective for the treatment of NSCLC.


Background
There has been a rapid increase in the use of medical technologies in the recent years. It is clear that proper use of these technologies can significantly improve the patients' conditions; however, uncontrolled and inappropriate use of them might lead to a waste of limited resources (1). PET scan is a non-invasive, expensive medical imaging technology that was introduced in 1950. This technology is currently used for diagnosis and treatment of various diseases, especially cancers across the world (2)(3)(4). Compared to the MRI and CT scan that explore the cancer lesions at an anatomical level, this technology can explore the cancer lesions at cellular and biochemical levels that might lead to a higher diagnostic performance (3). Various countries have commonly used PET scan for staging and treatment of cancers, especially lung cancers and non small cell lung carcinoma (NSCLC). Approximately 85% to 90% of lung cancers are non-small cell lung carcinoma (NSCLC) (2). Recent studies from the UK and Australia have shown that the technology is significantly cost-effective, particularly when it is used for staging and treatment of NSCLC and lymphoma (3,5). It is argued that the evidence regarding the technical performance and effectiveness of technologies are generalizable between the countries, but the cost-effectiveness evidence cannot be generalized to other countries (6).

Objectives
The main purpose of this research is to explore the cost effectiveness of PET scan in staging and treatment of NSCLC in Iran that might then help optimize the allocation of limited resources.

Materials and Methods
The process of using PET scan in staging lung cancer: PET scan is not normally used for screening and diagnosis of lung cancer. When a patient is diagnosed with lung cancer and the biopsy shows that the cancer is NSCLC, then normally a CT scan is performed to explore any metastasis (3). The accuracy of CT scan for detecting the cancer metastasis is moderate and performing a PET scan at this stage can improve the accuracy of staging (3). The accuracy of PET scan at this stage is different depending on the result of the CT scan, whether it is positive or negative. In addition, according to the marker studies (3,5), there are seven possible strategies for detecting metastasis in NSCLC after CT scan. Therefore, for the purpose of economic evaluation, we designed a standard decision tree model that included seven strategies after the CT scan was negative and seven strategies after the CT scan was positive consisting of a total of 14 strategies ( Table 1).
The data about technical and clinical performance of PET Those reports that met the quality criteria and were up to date and also had the potential for answering parts or all of the study questions were included. Iranian national databases including IranMedex and SID were searched to identify local and national socio-demographic data and data about the incidence of lung cancer and NSCLC, the staging of disease at the time of diagnosis and other local data required for the model. Retrieved studies and expert opinion were used to estimate the cost of each treatment strategy in Iran. The costs were divided into three categories including capital costs (depreciation costs of buildings and equipment), staff costs and other expenses (including cost of consumables, running and maintenance costs). The costs were estimated in both IR-Rials and US-Dollars with an exchange rate of 10.000 IR Rials per one US Dollar according to the exchange rate in 2008. Sensitivity analysis was used to explore any possible uncertainties in the socio-demographic and costing data. The data about the costs of each strategy were col-lected according to the local tariffs and expenses in Iran at the time of study. Sensitivity analysis was performed according to the following assumptions. These assumptions were made according to the current situation at the time of the study either from the robust evidence or expert opinions.

Results
The total costs of setting up a complete PET scan unit were 81600 million Rials (8.2 million US$). The total costs were 57000 million Rials (5.7 million US$) when we excluded the building costs and 37400 million Rials (3.7 million US$) when we excluded the building costs and the costs of the cyclotron unit (Tables 3, 4, 5 and 6). Assuming that a PET scan will perform about 3000 tests a year, the total annual cost of running a PET scan unit including building, equipment, staff and consumable products were 10860 to 13000 million Rials (1.08 to 1.3 million US$). Excluding the building costs, the total annual cost of running a PET scan were 9430 million to 11000 million Rials (0.9 to 1.1 million US$). Excluding the building costs and purchasing the FDG-18 from another cyclotron unit, the total annual cost of running a PET scan was 8850 to 10900 million Rials (0.8 to 1.09 million US$). Considering the above mentioned assumptions, the cost of a PET scan test was 3,000,000 to 4,400,000 Rials (300 to 440 US$). Assuming a discount rate of zero and a 10% increase in staff salary, the total projected costs of running a complete PET scan unit were about 28 trillion Rials (2800 million US$) after 20 years (Tables 3, 4, 5 and 6). Table 7 shows the minimum and maximum local costs of each single intervention that we used to estimate the costs of each strategy.
Option A -(low cost and low incidence of metastasis): Table 8 shows that when the CT scan was positive, the third strategy had the lowest cost per QALY followed by the first, fourth, seventh, fifth, sixth and second strategies. In addition, when the CT scan was negative, the first strategy had the lowest cost per QALY, followed by the third, seventh, fifth; fourth, sixth and second strategies. Option D -(high costs and the high incidence of metastasis): when the CT scan results were positive: the third strategy would be associated with the lowest cost per QALY and the fourth; sixth, fifth, second, seventh and the first strategy would be placed in the next steps in an increasing order. When the CT scan results were negative: the third strategy would be associated with the lowest cost per QALY and the fourth; sixth, fifth, seventh, first and second strategy would be placed in the next steps in an increasing order. The incremental cost effectiveness ratio (ICER) for each strategy was calculated for options A and D, presented in Tables 9 and 10.        (3,9) showed that when diagnosis of NSCLC was confirmed with conventional methods such as biopsy and CT scan, the use of PET-FDG in assessing the degree of malignancy before surgery was cost effective, especially in those patients who had a negative CT scan. The Scottish study indicated that the use of PET scan is effective only if the willingness to pay is higher than 60,000 pounds per QALY (9). Another study was published in 2003 in Australia showing that use of PET scan leads to an average increase of 0.046 years of patient life (about 17 days) and an average increase of $189 in cost per each patient. This means that the additional cost per additional QALY gained is about $ 41.087 (5). A third study was conducted in 2002 in Scotland that used a decision model with 5 strategies for the evaluation of PET. This study showed that using CT scan without PET scan leads to an increase in life expectancy and reduction in cost. But this approach leads to radiotherapy and unnecessary treatment in 36 percent of the patients. In contrast, when the technology alone and without FDG-PET scans was used, this figure reduced to 4%. In addition, it led to 0.7 increase in the average years and 236 pounds reduction in the cost per patient (10). Considering the small changes of this technology in the clinical and final outcomes (QALY) of the patients and its small savings costs, the final decision about using or not using this technology in Iran depends on whether the technical performance of the test is considered more important or the clinical final outcomes. If early and more accurate detection of the disease is more important than the final outcomes, then the technology should be considered useful. PET scan seems to be an effective technology and could be diffused in Iran based on a comprehensive technology rationing system at national and regional levels. To increase the efficiency and preventing induced demand, the indication for use of this technology in Iran should be NSCLC. Based on available information, some Iranian medical centers received the license for entering PET scan, but only two of them could enter this technology with attention to this point that PET scan has not been installed completely in these centers yet. The actual costs and clinical findings of PET scan were not available in Iran due to the unavailability of the technology, so the data were estimated based on international evidence. Drummond et al. argued that the clinical data can be used and generalized between the countries (6). To minimize the possible bias, we used the socio-demographic and costing data according to the local evidence available in Iran.