Lung cancer remains a significant public health challenge due to its high incidence and mortality, and despite advances in diagnosis and treatment, its survival rate remains low compared to other cancers.1 Low-dose CT screening has been shown to reduce mortality, as demonstrated in studies such as the NLST2 in the United States (US) and the NELSON study3 in Europe. However, the effectiveness of the inclusion criteria may vary by context. In Spain, for example, the USPSTF screening criteria (used in the US to establish the eligibility criteria in the US and like those on the NELSON study) would only detect 63.5% of lung cancer cases, with even lower detection rates in women.4 These disparities suggest the need to evaluate and potentially adapt international screening criteria to fit local contexts, taking also into account how these criteria may impact the feasibility of the program. This study aims to estimate the population eligible for a lung cancer screening program in Galicia (Spain) by applying international eligibility criteria and exploring different scenarios.

To estimate the eligible population, we used data from the 2018 Galician Risk Behavior Data System (SICRI) and the 2022 Municipal Voters Roll from the Instituto Galego de Estatística (IGE). SICRI-2022 is a telephone-based survey conducted on 7853 individuals aged 16 years and older, collecting detailed sociodemographic and smoking information. This dataset enabled us to estimate the prevalence of smokers and ex-smokers by health area, age group, and sex.

Eligibility for lung cancer screening was assessed using the 2013 and 2021 USPSTF guidelines, focusing on smoking in pack-years, the period of smoking abstinence (in ex-smokers). Additionally, three broad age ranges were considered, 50–80 years, 55–80 years and 65–80 years. These factors were combined into twelve scenarios (four for each age range). To determine the number of individuals meeting the smoking criteria by health area, sex, and age group, we estimated the proportion of daily smokers with more than 20 or 30 pack-years using data from the SICRI survey, applying these proportions to the population of Galicia using the Municipal Voters Roll. For ex-smokers, given that the SICRI survey did not collect daily consumption, we used current smokers’ data to calculate pack-years. ROC curves were then applied to estimate the minimum number of years of smoking required to meet the 20- or 30-pack-year criteria. This estimate was applied to the Galician population. Three different participation rates (40%, 60%, and 80%) were considered. We also estimated the required number of LDCT scanners to support the program. Each scanner was assumed to operate over 250 working days per year, with a scanning capacity of 40 patients per day (equivalent to 10,000 scans per year). Statistical analyses were conducted using Stata v.17.

The analysis showed significant variability in the number of subjects eligible for lung cancer screening in Galicia depending on the tobacco consumption and smoking cessation criteria applied, as well as the age range. Using the loosest screening criteria—defined as including individuals aged 50–80 years who are current or former smokers with more than 20 pack-years and, in the case of ex-smokers, less than 15 years of abstinence—a total of 249,099 persons across Galicia would be eligible for screening. In contrast, the strictest screening criteria—defined as including individuals aged 65–80 years who are current or former smokers with more than 30 pack-years and, for ex-smokers, less than 10 years of abstinence—reduce the eligible population to 53,931 individuals. These results were detailed by health area in Table 1.

With an assumed participation rate of 60%, the number of screened individuals would be approximately 149,459 under the loosest criteria and 32,358 under the strictest criteria. For example, under the loosest scenario, Vigo and A Coruña are projected to have 32,936 and 30,650 participants, while Pontevedra would have approximately 13,082 participants. Conversely, using the strictest criteria, Vigo and A Coruña would be reduced by 83% and 80% respectively, and Pontevedra by 70% (Table 2).

In terms of the number of low-dose CT scanners necessary to support the screening program, under the scenario using the loosest criteria (i.e., >20 pack-years and <15 years of abstinence for ex-smokers, and an age range of 50–80 years), the screening volume would require over three CT scanners in major regions—for instance, 3.1 scanners in A Coruña and 3.3 in Vigo. However, applying the strictest criteria (i.e., >30 pack-years and <10 years of abstinence for ex-smokers, with an age range of 65–80 years) would significantly reduce the demand for scanners, with estimates of 0.5 scanners for A Coruña and 0.7 for Vigo. Practically, under this strict scenario, A Coruña could manage its screening needs with one CT scanner operating at 20 tests per day over a full year or 40 tests per day over a six-month period. Overall, the total number of dedicated CT scanners needed by the Galician Health Service would vary based on the screening criteria chosen—from an estimated 14.9 scanners under the loosest criteria to 3.2 scanners under the strictest criteria.

This study is the first to estimate the number of lung cancer screening candidates coupled with scanner needs in a specific region using real data and a bottom-up strategy. The findings reveal that the number of eligible individuals varies significantly based on screening criteria such as smoking history and years of abstinence. Stricter criteria improve program effectiveness but reduce the number of detected cases, while looser criteria increase cases but also costs and false positives.5

Our results show that shifting from broad to strict criteria can reduce the number of candidates fivefold. These estimates are crucial for planning screening implementation, particularly in countries with universal healthcare, where resource allocation must not disrupt existing services. A key challenge is the availability of CT scanners and radiologists.

Our study benefits from reliable population data in a setting with universal healthcare coverage, ensuring that estimates reflect the entire eligible population. However, there are several limitations. First, the study lacks a detailed analysis on false positives and the subsequent resource implications of incidental findings, which are known to account for 20–50% of screening outcomes.6–9 Second, the inclusion criteria did not include lung cancer risk factors other than tobacco use and age. Third, although the American Cancer Society recently recommended removing the years-since-quit threshold, aligning with the NELSON trial, which did not set such a limit. In Spain, this change would be relevant, as a national study found that 34.1% of diagnosed lung cancer cases would have been excluded due to quitting more than 15 years ago.4 However, this study follows the criteria used in major European clinical trials and pilots, assuming Spain would adopt similar guidelines for implementation. Lastly, the estimates for CT scanners were based on clinician-reported data. However, the lack of data on existing scanners makes it not possible to assess the gap between required and available resources.

In conclusion, healthcare authorities should carefully decide on the inclusion criteria and allocate sufficient material and human resources before implementing lung cancer screening programs, while also integrating effective anti-smoking interventions for active smokers within these initiatives. It is worth mentioning that most lung cancer screening research is based on clinical trials rather than real-world data. Assessing real-life outcomes is crucial to refine initial predictions before expanding screening to the entire target population.