Calculation from absolute risk reduction

Risk Reduction - The Core Concepts

• Control Event Rate (CER): Risk of an outcome in the placebo or control arm.
  • Formula: $CER = \frac{\text{Events in control}}{\text{Total in control}}$
• Experimental Event Rate (EER): Risk of an outcome in the treatment or experimental arm.
  • Formula: $EER = \frac{\text{Events in treatment}}{\text{Total in treatment}}$
• Absolute Risk Reduction (ARR): The absolute difference in event rates between control and treatment groups; the most clinically relevant measure.
  • Formula: $ARR = CER - EER$
• Relative Risk Reduction (RRR): The proportional reduction in risk; often used in drug advertising as it can appear more impressive than ARR.
  • Formula: $RRR = \frac{ARR}{CER} = \frac{CER - EER}{CER}$

⭐ High-Yield: RRR can be misleading. A large RRR may hide a small ARR if the baseline risk (CER) is very low. Always base clinical decisions on absolute risk reduction (ARR).

The Calculation - From ARR to NNT

The Number Needed to Treat (NNT) is the reciprocal of the Absolute Risk Reduction (ARR). It quantifies how many patients need an intervention for one to benefit. A lower NNT is better.

• Step 1: Find the Absolute Risk Reduction (ARR)

  • This measures the true difference in risk.
  • $ARR = CER (Control Event Rate) - EER (Experimental Event Rate)$

• Step 2: Calculate the NNT

  • This is the inverse of the ARR.
  • $NNT = 1 / ARR$

For adverse outcomes, the same logic applies to the Number Needed to Harm (NNH) and Absolute Risk Increase (ARI).

• $ARI = EER - CER$
• $NNH = 1 / ARI$

⭐ An NNT of 1 is ideal, meaning every patient treated sees a benefit. Always round the calculated NNT value up to the next whole number, as you cannot treat a fraction of a patient.

Clinical Interpretation - So What?

• NNT (Number Needed to Treat): Average number of patients who need treatment to prevent one additional bad outcome.

  • A lower NNT is better.
  • An ideal NNT is 1, where every patient treated benefits.
  • Context is key: An NNT of 20 for preventing death is excellent, but for preventing a mild headache, it's poor.

• NNH (Number Needed to Harm): Average number of patients who need treatment to cause one additional harmful outcome.

  • A higher NNH is better.
  • An ideal NNH is infinite (the intervention causes no harm).

• Clinical Decision-Making:

  • A clinically desirable intervention has a low NNT and a high NNH.
  • The balance depends on the severity of the benefit vs. the harm. Is preventing a stroke worth the risk of minor side effects?

⭐ NNT and NNH are always rounded up to the next whole number, because you cannot treat a fraction of a person.

High‑Yield Points - ⚡ Biggest Takeaways

• The Number Needed to Treat (NNT) is the reciprocal of the Absolute Risk Reduction (ARR), so the formula is NNT = 1 / ARR.
• Similarly, the Number Needed to Harm (NNH) is the reciprocal of the Absolute Risk Increase (ARI), calculated as NNH = 1 / ARI.
• ARR itself is the control event rate (CER) minus the experimental event rate (EER).
• A lower NNT indicates a more effective and potent intervention.
• Always round NNT and NNH up to the next whole number.