Lung-RADS or Fleischner? Picking the Right Pulmonary Nodule Pathway

Two radiologists can look at the same 7 mm solid pulmonary nodule and recommend different follow-up, both correctly. The difference is not judgment — it is which system applies. Lung-RADS and the Fleischner Society criteria are not competing guidelines, and choosing between them is not a matter of preference.

Getting this wrong is common, and it is the kind of error that survives peer review because both recommendations look defensible in isolation.

The one question that decides it

Ask: was this CT performed as part of a formal lung cancer screening programme?

If yes, use Lung-RADS. If no, use Fleischner.

That is the whole rule. Not the patient's smoking history, not the nodule's appearance, not whether the patient would have qualified for screening. The question is what the exam was ordered as.

• A low-dose CT ordered as annual lung cancer screening in an eligible patient → Lung-RADS.
• A nodule seen incidentally on a CT pulmonary angiogram, a staging CT, a trauma pan-scan, a CT abdomen that caught the lung bases → Fleischner.

A heavy smoker who has never been enrolled in a screening programme, presenting with chest pain and getting a CTPA, has an incidental nodule. Fleischner applies, and the patient's risk factors feed into Fleischner's own high-risk branch — not into Lung-RADS.

Why the distinction exists

Screening and incidental detection have different pre-test probabilities and different downstream infrastructure.

Lung-RADS was built for a population that has been deliberately selected for elevated risk, is being imaged on a standardised low-dose protocol, and is enrolled in a programme with tracking and navigation. It produces a category (1 through 4X, plus modifiers) that drives a management recommendation and, in the United States, is tied to reporting requirements for screening programmes.

Fleischner was built for the messier reality of a nodule that turned up while you were looking for something else, on a scan with no protocol standardisation, in a patient whose risk you have to estimate yourself. It produces a follow-up interval, not a category.

Because the populations differ, the size thresholds differ. A nodule that is unremarkable in one system can be actionable in the other.

Where the thresholds diverge

The most-cited divergence is at the small end. Under Fleischner, a solitary solid nodule under 6 mm in a low-risk patient needs no routine follow-up at all. Under Lung-RADS, a solid nodule under 6 mm at baseline is category 2 — benign appearance or behaviour — and the patient still returns in 12 months, because they are in an annual screening programme regardless.

That is not a contradiction. The screening patient was always coming back. The incidental patient was not, and Fleischner is deciding whether to create a follow-up pathway that would not otherwise exist.

Similar logic explains the ground-glass thresholds. Lung-RADS tolerates substantially larger pure ground-glass nodules before escalating, because the screening population's ground-glass nodules are overwhelmingly indolent and the programme will see them again next year anyway.

Four mistakes worth watching for

1. Applying Fleischner to a screening exam. This is the most consequential direction of error. It can under-call a screening finding that the programme's own protocol would have escalated, and in a formal programme it also breaks the reporting the programme depends on.

2. Applying Lung-RADS to an incidental nodule. Less dangerous, but it produces a category that means nothing outside a screening context and can trigger follow-up the patient does not need. It also implies the patient is enrolled in something they are not.

3. Applying Fleischner where nothing applies. Fleischner explicitly excludes several groups: patients under 35, patients with known or suspected primary malignancy, and immunocompromised patients. A nodule in a patient with a known primary is a potential metastasis and is managed on oncological grounds, not by a size-based follow-up table. A nodule in an immunocompromised patient is a potential infection and may need attention in days, not months.

4. Measuring inconsistently. Fleischner 2017 uses the mean of the long-axis and short-axis diameters on the same section, rounded to the nearest millimetre — not the single longest dimension. A 7 × 5 mm nodule is a 6 mm nodule. Reporting it as 7 mm can move it across a threshold. If you are going to use a size-banded system, the measurement convention is part of the system.

Reporting it so the referrer can act

Whichever system applies, name it. "Recommend CT at 6-12 months" is an instruction. "6 mm solid nodule. Fleischner Society 2017 criteria for an incidental nodule in a low-risk patient: recommend CT at 6-12 months, then consider CT at 18-24 months" is a defensible recommendation that a referring clinician can act on and a future reader can audit.

State the version too. Fleischner 2017 differs from the 2005 and 2013 iterations, and Lung-RADS has moved through several revisions. A recommendation without a version is hard to check later.

Get the interval without the table

Our free Fleischner Society calculator takes nodule type, size, multiplicity and risk category and returns the 2017 recommendation, with the full decision table on the page so you can see which branch you landed on. No sign-up.

For the screening side, the Lung-RADS v2022 calculator does the same thing for a screening LDCT, including the lower thresholds that apply to a nodule new on an annual round.

If you would rather have the recommendation written into the report itself — with the guideline named and the version cited — that is what MyRadAgent does while you dictate.

The short version

Screening exam, screening system. Incidental nodule, Fleischner. Check the exclusions before you reach for the table, and measure the way the table expects. Most disagreements about nodule follow-up turn out to be disagreements about which question was being asked.