We believe that the conclusions of the chapter “Early diagnosis of lung cancer. The future of screening”, in the recently published SEPAR monograph,1 which claim that the absence of a lung cancer screening program in Spain is more a question of logistics than money, merit a special comment.

While we agree with some of the statements, we disagree with many others. For example, no reference is made to the barely perceptible increase in the number of early-stage tumors detected in the NLST.2 Cases diagnosed at stage I did not increase more than 10% when the third screening round was compared round with the first, and moreover, in all rounds, the proportion of cases detected at stage III and IV remained steady at 30%. Detecting disease at an earlier stage is the cornerstone of screening programs, since it indicates a change in the clinical course of the disease. The high percentage of detection of lung cancer at advanced stages indicates that screening is not effective in a significant proportion of the subjects analyzed.

In our opinion, this understates the risk induced by radiation, since the proposed screening programs involve an annual LDCT. The authors point out correctly that in each round there are approximately 25% positives. False positives can be reduced by using volumetric criteria instead of diameter, as indicated by the NELSON study. A positive result involves performing at least one diagnostic CT (in addition to the follow-up CTs required if subcentimeter nodules are detected), which considerably increases exposure to radiation. Approximately every 4 years, participants would receive at least 12mSv (4 LDCT+standard CT), which between the ages of 55 and 80 years would add up to a total of 72mSv. It has been calculated that participants in a screening program would receive more radiation than atomic bomb survivors or nuclear plant workers (up to 280mSv in the most conservative estimate).3

Lung cancer screening is compared with breast and colorectal cancer screening, but they are not comparable. Unlike lung cancer screening, which is selective (smokers or former smokers), breast and colorectal screening programs are population programs, in which a positive finding on mammography or occult blood in stool can generally be confirmed in a matter of days by image-guided biopsy or colonoscopy with biopsy. A positive finding in LC screening of a subcentimeter nodule might persist for months or years before growth is confirmed or ruled out and subsequent actions are planned.

No European study has proved the effectiveness of screening. Additionally, a recent study using NLST data4 suggested that 2-yearly screening would be just as effective as annual screening, and the accompanying editorial even claimed that the annual screening interval is not based on biological evidence, but rather on a pragmatic decision based on organizational considerations. Breast cancer screening is performed every 2 years. Data are available in Spain on the organizational aspects of a screening program.5 Up to 1,700,000 individuals meet the screening criteria, and approximately 162 scanners dedicated exclusively to screening would be needed. The scanners alone would cost close to €1 bn.

While economic and organizational aspects are relevant, we firmly believe that the risk–benefit balance in the screening of lung cancer with LDCT is still questionable. Although screening might reduce lung cancer deaths from 21 to 18 individuals for each 1000 screened, the level of iatrogenic complications would be high. In our opinion, scientific societies should debate these issues before presenting them in a text which might be taken as recommendations, but which, in this case, do not have sufficient consensus.