Lung cancer (LC) is a malignant growth of cells in the lung or bronchial system. The symptoms of LC are nonspecific and usually appear when the disease has already spread, making early diagnosis difficult. LC is the fourth most common type of cancer in Spain and, in terms of mortality, it is the leading cause of cancer death, responsible for more than 23,000 deaths in 2023 (20.6% of total cancer mortality).1 Although there are several risk factors of developing LC, tobacco addiction is by far the most important preventable cause of LC, responsible for 90% of cases in men and 80% in women.2 Passive smoking or exposure to tobacco smoke in non-smokers increases the risk by 20-30% compared to non-smokers not exposed to passive smoke.3

A screening test that detects the disease at an early stage would help modify the clinical course of the disease, facilitating early diagnosis and treatment, thereby improving survival by preventing disease progression. Imaging techniques are used for screening, with low-dose radiation computed tomography (LDCT) being the only test currently recommended. Overall, the LC screening with LDCT may have a benefit in LC-specific mortality compared to no screening, but the benefit in overall mortality and morbidity is inconclusive.4 In economic terms, many studies conclude that screening with LDCT is more expensive and more effective than no screening, but it could be cost-effective only under certain conditions.4,5 The use of blood or exhaled air biomarker panels is also proposed, although they are in an early phase of development and validation.

Early detection of LC using LDCT is increasingly being implemented as a population screening program, although there is still some uncertainty about its risk-benefit ratio and cost-effectiveness. In this context, a budget impact analysis was carried out to 1) estimate the cost of its implementation in high-risk population for the National Health System (NHS) in Spain, and 2) provide useful methods for other researchers and decision-makers considering the implementation of LC screening in their healthcare settings.

The analysis performed allowed us to estimate not only the net budget impact of implementing a LC screening program in Spain, but also the potential number of people diagnosed by cancer stage in each scenario (Table B1 of the Appendix B). Overall, the screening program, especially the annual one, could diagnose more people than the current scenario without screening. In addition, 40-56% (annual screening) and 32-50% (biennial screening) of diagnoses are detected at stage I, whereas in the no-screening scenario, 52% of diagnoses are detected at stage IV. Notably, biennial screening would detect almost the same number of stages I (32%) and IV (35%) cancers when the age interval is specified as 55-65 years.

The gross budget impact for the current scenario (without screening) is estimated at €1708.19 million in five years (Table 1). Meanwhile, the budget impact of the implementation of a LC screening program varies depending on the established age range delimiting the target population. Among the hypothetical scenarios (with screening), the total gross budget impact can range between €3737.17 million (biennial screening, 55-65 years) and €10 009.54 million (annual screening, 50-80 years), resulting in a net budget impact between 2028.98 million and 8301.35 million of additional Euros in the first five years of the implementation (Table 1).

Disaggregating the gross budget impact by cost groups reveals higher expenditures across all categories in the hypothetical scenarios (Tables B2 and B3 of the Appendix B). The costs of treatment and follow-up represent the largest item (81% of the total impact) in the scenario without screening, while in the hypothetical scenarios the highest costs are those classified as screening costs (44%-62% of the total gross budget impact) (Table 2). If we do not consider the latter in the screening scenarios, the weight of the treatment and follow-up costs on the total impact is lower in the scenarios with screening than in the scenario without screening (56%-70% vs. 81%, respectively) (Table 2). Additionally, within the screening cost group, LDCT and nursing consultation components show the higher percentages (i.e., 53–59% and 14–21% of total screening costs, respectively) (Tables B2 and B3 of the Appendix B).

Regarding investment in CT scanners, the implementation of the screening requires a large investment in new equipment (Table 3). Assuming the acquisition of 100% of the necessary equips, the investment can range from €314.72 million and €781.85 million (screening between 55-65 years old (the shortest age range) and 50-80 years old (the widest age range), respectively) in the annual screening. In the case of biennial screening, the investment can vary between €170.7 million and €417.6 million (for the shortest and the widest age range, respectively). Also, given that the implementation would be progressive, the investment in the first year would be the highest in relative terms (approximately 22% of the screening costs of that year, both in annual and biennial programs; results not shown).

The net budget impact by regions (Table 4) would be greater in those regions where the population susceptible to be screened is larger (i.e., Andalusia, Catalonia and Madrid), assuming equal costs for the entire national territory. However, this impact could be undervalued in the regions that, with a smaller target population, could not obtain volume discounts for CT scanner purchase.

The sensitivity analysis shows how, in general, the most relevant variations in the net budget impacts are observed when the response rate and the participation rate are modified (Figures C1 and C2 of the Appendix C). Also, when the percentage of positive LDCT increases from 2.3% to 5.7% (value estimated from Field et al.16), generating an increase of 48%-50% of the net impacts.

There is a large uncertainty surrounding the necessary number of CT equipment to be acquired. However, varying this parameter would not significantly affect the net budget impact. In fact, it could range between €1.9 and €7.9 billion, considering all screening scenarios (Figures C1 and C2 of the Appendix C). Specifically in the case of not acquiring any CT equipment, the net impact of an annual program would be €3.2 billion for the shortest age range (55-65 years) and €7.5 billion for the widest range (50-80 years), while the net impact of a biennial program would be €1.9 billion or €4.3 billion, respectively.

In healthcare decision-making, a budget impact analysis offers practical short- and medium-term budget impact projections that are crucial for resource planning. These projections should be considered alongside cost-effectiveness results when evaluating the introduction of a new health technology into the NHS.

Here, we present such projections in the context of a population-based LC screening program by evaluating several hypothetical scenarios based on different age ranges and screening frequencies. Our findings indicates that, regardless the hypothetical scenario considered, the program would entail a substantial cost for the NHS, consistent with the conclusions of Ruano-Ravina et al.13

The own costs generated by the screening program (i.e., screening costs in Table A1 of the Appendix A) represent the largest contribution to the total gross budget impact. Within this cost group, although the LDCT and nursing consultations are the main cost components, accounting for over 70% of the share, the investment in CT scan equipment constitutes a large part: around 11%-12% of the screening costs are due to the acquisition of new equipment (assuming the purchase of 100% of the equipment needed). Reducing the number of CT scanners purchased would not greatly affect the net impact: it may be reduced by only 5%-9% if the share of CT scanners purchased is reduced to 50% or 0%, respectively; however, it is noteworthy that given the estimated useful life of the CT equipment, this investment (if necessary) would need to be repeated every 10 years.17

The results also show how implementing a LC screening program could reduce the weight of the treatment and follow-up costs on the total budget impact, which is in line with previous international results.18 This may be explained by the lower need for treatments for advanced LC, which are more expensive than those for early stages. The detection of LC in early stages by means of screening program (Table B1 of the Appendix B) could improve the prognosis of the disease by avoiding or delaying the onset of advanced stages.

To the best of our knowledge, this is the first budget impact analysis of the LC screening in Spain comparing 16 hypothetical scenarios with screening. Internationally, Silvestrini et al.19 estimated a lower incremental five-year impact for a biennial screening for individuals aged 55-74 in Argentina (approximately $129 million), which may be explained by lower unit costs and participations rates, for example. Our estimates are more in line with those from the United States, where annual screening costs exceed $1.3 billion.20,21 Both Goulart et al.20 and us determined similar annual costs of the screening program: $1.30 billion for the US, and €1 billion for Spain. To be comparable, this cost for Spain (not included in the results of this paper) refers to the cost of year 1 of an annual screening program for a target population of 55-75 years, with a 50% participation rate and 0% of CT equipment purchased (Goulart et al.20 set a target population of 55-74 years and a one-year horizon). Despite the possible differences in costs, population characteristics and healthcare system between Spain and the US, the results of the impact of the implementation of a national screening converge in economic figures in both studies.

Our analysis has several limitations. First, the screened population is obtained from some assumptions: 1) the response rate to the screening invitation was based on assumptions; 2) the percentage of high-risk population of developing LC was obtained through the SNHS 2011-12, because it is the last national survey with the data for this calculation; 3) the participation rate was estimated based on the participation in the colorectal cancer screening program in Spain; and 4) the distributions of LC by stage and treatment were calculated from international studies,12,22 given the scarcity of data for the Spanish population. Additionally, only first-line treatments are considered.

Second, due to the lack of studies for Spain, the effectiveness of LDCT was extracted from the international literature11,12. Additionally, data from de Koning et al.11 refer only to a male population, while the population included in our analysis is made up of both sexes.

Third, the real capacity of the available CT scanners in the Spanish NHS to cover the demand generated by the LC screening program was unknown, so the acquisition of 100% of the necessary CT scanners was assumed. The fewer CT scanners needed, the lower budget impact might be. In contrast, the current budget impact would be underestimated if the CT equipment were not working at full capacity, as we assumed. Also, the costs of maintenance and replacement of CT scanner parts, which were not included in the analysis, will have to be assumed by the NHS. Furthermore, the unit cost of a CT scanner was obtained from the literature.13

And fourth, the cost of preventive public interventions (e.g., smoking cessation programmes), which is recommended to be implemented in conjunction with LC screening, were not considered. Their consideration could have a double effect on the gross budget impact of the screening scenarios: 1) it would increase the budget impact by adding the costs of their application, and 2) if these actions were effective, the budget impact could decrease by reducing the target population in long term (smokers and ex-smokers who have been quitting in the last 15 years). Likewise, the cost of potential public campaigns aimed at raising awareness of the benefits of participating in screening programs should be added. Finally, the budget impact would be increased if the costs derived from the diagnosis and treatment of diseases other than LC found incidentally during screening were included.

The results of the budget impact analysis estimated for one organization are not generalizable to another organization. However, the methods applied and described in this article have followed internationally established methods7 and, therefore, may be helpful for other researchers or economic managers to perform similar analyses when considering implementing LC screening in their settings or countries, especially for European countries with national cancer screening programs and public health systems comparable to the Spanish one. In addition, the greatest utility of this article is to make available to the Spanish regional authorities an instrument adaptable to each of their regions so that, when the time comes to plan the implementation of the screening program, they can make their updated estimations with the data of each organization.

In conclusion, and according to the results, implementing a LC screening program in Spain would generate a high cost for the NHS, amounting to between €2 and €8.3 billion more than the current scenario without screening, depends on the characteristics of the screening. The non-acquisition of any CT scanner could reduce the net impact to between €1.9 and €7.5 billion.

This study does not rely on a specific database. All data inputs used in the budget impact model were obtained from available sources, including published literature and national statistics, which are duly cited throughout the manuscript. Therefore, all relevant materials needed to replicate the model are accessible through the referenced publications.

The corresponding author, on behalf of the other authors guarantee the accuracy, transparency and honesty of the data and information contained in the study, that no relevant information has been omitted and that all discrepancies between authors have been adequately resolved and described.

A. Hernández-Yumar participated in the design, acquisition, analysis, and interpretation of data, as well as drafting the work. C. Valcárcel-Nazco and L. García-Pérez participated in the design, acquisition, analysis, and interpretation of data, and critically reviewed the work. P. Cantero-Muñoz critically reviewed the work. L. García-Pérez and P. Cantero-Muñoz contributed to project administration. All authors read and approved the final manuscript.

The authors would like to thank Benjamín Rodríguez Díaz for his valuable assistance with the search and identification of unit cost data.

This study was commissioned and funded by the Spanish Ministry of Health in the framework of activities carried out by the Spanish Network of Agencies for Assessing National Health System Technologies and Performance (RedETS). The study was conducted independently of study sponsors. There was no sponsor involvement in the study design; collection, analysis and interpretation of the data; in writing of the manuscript; or in the decision to submit the manuscript for publication.

None.