How to Read an ACC/AHA Guideline Recommendation

Lesson 4 Supplement — Cardiovascular Pharmacology

Download PDF

How to Read an ACC/AHA Guideline Recommendation

Lesson 4 Supplement — Cardiovascular Pharmacology

Every cardiology guideline from the ACC and AHA rates each recommendation on two independent dimensions: how strongly they recommend it (Class of Recommendation, COR) and how good the evidence is (Level of Evidence, LOE). Both appear every time — together, they tell you how much weight to give a recommendation in clinical practice.

Part 1 — Class of Recommendation (COR): “How Strongly?”

The COR tells you whether you should do something, consider doing it, or avoid it altogether.

Class Plain Language What It Means Clinically
Class I Should do this Strong evidence or agreement that benefit outweighs risk. Do it unless contraindicated.
Class IIa Reasonable to do Weight of evidence favors benefit. Most patients will benefit — sound choice in the right patient.
Class IIb May consider this Benefit less certain. Evidence is weaker or conflicting. Use clinical judgment.
Class III: No Benefit Don’t bother Evidence shows no benefit in this specific indication. Doing it is not harmful — but it doesn’t help.
Class III: Harm Don’t do this Evidence of harm. Doing it in this situation is actively dangerous.

The language cue: Guideline wording is not accidental. “Is recommended” = Class I. “Is reasonable” = Class IIa. “May be considered” = Class IIb. “Should not be done” = Class III. Once you know the translation, you can read the class without looking it up.

Part 2 — Level of Evidence (LOE): “How Good Is the Data?”

The LOE tells you what kind of research supports the recommendation. A weak LOE does not automatically weaken the recommendation — it may simply mean that a large RCT has never been done or is not feasible.

Level What It Means
A Multiple high-quality RCTs, or meta-analyses of high-quality RCTs. The gold standard.
B-R At least one moderate-quality RCT, or meta-analyses of moderate-quality RCTs. Solid randomized data, more limited scope.
B-NR Well-designed nonrandomized studies, observational data, or registry studies. Good data, but not randomized.
C-LD Randomized or nonrandomized studies with significant design limitations, or mechanistic/physiological studies.
C-EO Expert opinion based on clinical experience. No direct trial data exists — this is the panel’s best collective judgment.

Context matters: Many Class I recommendations carry Level C-EO — particularly for procedures where a placebo-controlled trial would be unethical. That does not make them optional. “Shock a patient in pulseless VF” is Class I, Level C-EO because nobody ran a randomized trial.

Part 3 — Reading a Recommendation

“High-intensity statin therapy is recommended for all patients with clinical ASCVD unless contraindicated (Class I, Level A).
  • Class I — You should do this. Strong recommendation.
  • Level A — Multiple high-quality RCTs support it (JUPITER, 4S, HPS, ASCOT-LLA, CARDS, etc.).
  • “Unless contraindicated” — The only acceptable reason not to follow Class I is a specific contraindication.
“Aldosterone antagonists may be considered in patients with HFpEF to decrease hospitalizations (Class IIb, Level B-R).
  • Class IIb — May be considered. Weaker recommendation — benefit is not clearly established.
  • Level B-R — One moderate-quality RCT (TOPCAT) with limitations. Real data, but not definitive.
Clinical Rule

Class I + Level A is the gold standard of guideline-based care. If a patient with a Class I, Level A indication is not on that therapy, you need a documented reason why.

This is one of 13 free reference sheets from the APP Cardiology Academy — no account required.

Browse all resources See the full curriculum