OpenOnco
UA EN
← All news
July 20, 2026 guidelines evidence

What NCCN category 2A actually means

2A is widely read as "weak evidence, weak recommendation." It is not. The panel agreement behind 2A is the same as behind category 1 — and 2A is what you get by default when no category is printed at all.

Background reading, not clinical guidance. This describes published research; it is not a recommendation for any individual case, and the cited studies may have been superseded.

NCCN grades each recommendation on a scale that combines two separate things: how strong the underlying evidence is, and how much the panel agreed. Reading the label as a single measure of confidence is where the misunderstanding starts.

The part people get wrong

Category 1 and category 2A require the same level of panel agreement — NCCN's "uniform consensus," defined as at least 85% support. They differ only in the evidence beneath them: category 1 rests on high-level evidence such as randomised phase 3 trials, 2A on lower-level evidence.

So 2A does not mean the panel was lukewarm. It means the panel was just as united, on a thinner evidence base.

The categories that do encode disagreement are further down: 2B is lower-level evidence with only 50–85% panel support, and category 3 means real dispute, with at least a quarter of the panel objecting.

2A is also the default

Every NCCN guideline carries a standing line: *all recommendations are category 2A unless otherwise indicated*. A recommendation with no visible category is not unclassified — it is 2A.

How much of the guidelines this covers

An independent academic analysis — not an NCCN publication — reviewed the 2019 guidelines for the ten most common cancers in the United States and counted 1,818 staging, treatment and surveillance recommendations. Of those, 7% were category 1, 87% were 2A, 6% were 2B, and none were category 3.

The same authors had counted 1,023 recommendations in 2010 and found essentially the same distribution. The guidelines grew by 77% over the decade; the proportion resting on high-level evidence barely moved.

Two caveats on that number. It covers roughly ten of NCCN's eighty-plus guidelines and excludes screening and supportive care, so it is not a figure for NCCN as a whole. And it is a count of recommendations, which says nothing about how often each one is actually used.

Why this matters for reading a plan

Most of oncology practice runs on expert consensus rather than phase 3 trials, and the guidelines say so openly if you read the categories correctly. That is not a scandal — much of medicine could never be randomised — but it does change what "guideline-recommended" is evidence of.

OpenOnco surfaces the category alongside each cited recommendation for this reason. A clinician should be able to see whether they are looking at a randomised result or a room full of experts who agreed.

Sources

  1. Desai AP, Go RS, Poonacha TK. Category of evidence and consensus underlying National Comprehensive Cancer Network guidelines: Is there evidence of progress? International Journal of Cancer, 2021;148(2):429-436 · 32674225
  2. Poonacha TK, Go RS. Level of scientific evidence underlying recommendations arising from the National Comprehensive Cancer Network clinical practice guidelines. Journal of Clinical Oncology, 2011;29(2):186-191 · 21149653
  3. National Comprehensive Cancer Network. NCCN Categories of Evidence and Consensus · https://www.nccn.org/guidelines/category_1

Comments

Threads are public and indexed — please keep personal medical data out of them. Found a clinical error? Open an issue instead.

Comments are not switched on yet — the backend still needs to be configured.