Who should have a low-dose CT scan for lung cancer when they feel well?

Australia now has a national answer. The National Lung Cancer Screening Program offers free screening every two years to people aged 50 to 70 who have no symptoms suggestive of lung cancer, currently smoke or quit within the past 10 years, and have smoked at least 30 pack-years.

That narrow entry point is deliberate. Low-dose computed tomography, or CT, reduced deaths from lung cancer in randomised trials of people at high tobacco-related risk. It also found nodules that were not cancer, led to extra scans and invasive tests, and detected some cancers that may never have caused illness. Screening works best when the chance of benefit is high enough to justify that cascade.

The program starts with risk, not a scan request

A pack-year combines the amount and duration of cigarette smoking. One pack of 20 cigarettes a day for 30 years is 30 pack-years; two packs a day for 15 years is also 30. The calculation is imperfect, especially when smoking patterns have changed, so eligibility is confirmed with a requesting practitioner.

All four Australian criteria have a purpose. Age and smoking history identify a group with a higher starting risk. The 10-year limit after quitting keeps the program closer to the populations in which benefit was tested. The absence of symptoms separates screening from diagnosis.

That last distinction can be missed. Screening asks whether an apparently well person in a defined risk group has an early cancer. A new persistent cough, coughing blood, unexplained weight loss or another concerning symptom poses a diagnostic question. Someone with symptoms needs clinical assessment based on the symptom and context, not reassurance from eligibility for a future screening round.

What the mortality result actually showed

The NELSON trial randomised 15,792 current or former smokers aged 50 to 74 in the Netherlands and Belgium. The screening group had low-dose CT at baseline, then at years 1, 3 and 5.5. The control group had no screening. Follow-up continued for at least 10 years.

Among the 13,195 men in the primary analysis, lung-cancer deaths occurred at 2.50 per 1,000 person-years with screening and 3.30 per 1,000 person-years without screening. The cumulative rate ratio was 0.76, with a 95% confidence interval from 0.61 to 0.94. Lung-cancer incidence was higher with screening—5.58 versus 4.91 cases per 1,000 person-years—because CT found more cancers.

The trial did not show fewer deaths from every cause. All-cause mortality was 13.93 versus 13.76 per 1,000 person-years, a rate ratio of 1.01. The women’s subgroup was much smaller, so its estimate was less precise.

A 2022 Cochrane review pooled 11 trials with 94,445 participants. Eight trials contributed to the mortality analysis. Low-dose CT reduced lung-cancer mortality by 21%, with a risk ratio of 0.79 and confidence interval from 0.72 to 0.87. The authors estimated that 226 high-risk people would need screening to prevent one lung-cancer death.

Those figures support a program for selected high-risk adults. They do not support adding CT to a low-risk “longevity scan” menu. When starting risk falls, many more people must be scanned to prevent one death, while false alarms, radiation and incidental findings continue.

A nodule begins classification, not diagnosis

Low-dose CT produces detailed cross-sectional images using less radiation than a standard diagnostic chest CT. It can reveal small lung nodules: rounded areas that may represent old infection, inflammation, scar tissue or cancer. The scan alone often cannot settle which explanation is correct.

NELSON used nodule volume and growth over time to decide what happened next. Across the screening group, 9.2% had at least one additional CT after an initially indeterminate result, while 2.1% were referred for a suspicious nodule. Serial comparison helped avoid sending every small nodule directly to biopsy.

Australia’s program also uses a structured nodule-management protocol. A very-low-risk result usually returns to screening in two years. Other findings may lead to another CT in 3, 6 or 12 months. A very-high-risk result prompts specialist referral. The pathway is designed around probability and change, rather than treating every spot as cancer.

This is why previous images and reliable follow-up are part of the intervention. The benefit in the trials did not come from owning one CT image. It came from repeated screening, consistent interpretation and timely investigation when a finding crossed a risk threshold.

More detection carries a cost

The US National Lung Screening Trial, or NLST, randomised 53,454 high-risk adults to three annual rounds of low-dose CT or chest X-ray. Lung-cancer mortality was 247 versus 309 deaths per 100,000 person-years, a 20% relative reduction. Under the trial’s older definition, 24.2% of CT screening rounds were positive and 96.4% of those positive results were false positives.

That 96.4% is not a forecast for Australia. NLST used a size-based rule that labelled many nodules positive, while current protocols use more structured risk and growth assessment. It does show why the definition of an abnormal result can change the number of people sent for further testing.

Across the Cochrane review, invasive tests were 2.6 times as frequent with low-dose CT screening. The estimated proportion of screen-detected cancers considered overdiagnosed was 18%, but the confidence interval ranged from 0 to 36% and the evidence was low certainty. Overdiagnosis means finding a genuine cancer that would not have caused symptoms or shortened life. It cannot usually be identified with certainty in one person at the time of detection.

The useful conversation happens before enrolment

Australia’s program places shared decision-making before the first scan. The clinician confirms smoking history, checks whether the person can undergo the scan and any necessary investigation, explains possible follow-up, and offers smoking-cessation support. A screening program does not replace reducing ongoing tobacco exposure; the two approaches address different parts of risk.

For an eligible person, a concrete question is: am I prepared for the full pathway if the scan finds something uncertain? For someone outside the criteria, the question is different: what is my actual lung-cancer risk, and would a scan outside the program have evidence that its benefits exceed its harms?

The central result is not that CT can see small nodules. It is that repeated, risk-targeted low-dose CT reduced lung-cancer deaths in selected current and former smokers. Australia has turned that result into a governed pathway for people without symptoms—complete with eligibility, structured reporting, recall and follow-up.

Sources

Evidence note: This Insight is general education, not individual medical advice or a recommendation to request imaging. Screening eligibility, symptoms and follow-up decisions require assessment by a qualified clinician.