Number Needed to Treat (NNT): How to Calculate and Interpret

How to easily calculate and retrieve useful information from survival curves using absolute risk reduction and number needed to treat (NNT).
This is a practical and straightforward biostatistics lecture focused on the estimation of absolute treatment effect. Treatment effect is usually assessed on a relative scale (through the hazard ratio) – the Kaplan Meier survival curves being only illustrative. The number needed to treat (NNT) is sometimes missing from the publications. We explicit the background supporting a greater use of the absolute risk reduction to assess treatment effect as well as a practical and straightforward method to assess absolute treatment effect from Kaplan Meier survival curves.
This lecture is targeted to a broad audience including clinicians who want to extract insightful and clinically relevant information from the medical literature. This video could be efficiently used in the setting of evidence-based medicine journal club.
The learning objectives of this course are:
- Understand the methodological and clinical relevance of absolute treatment effect evaluation
- Know how to calculate the number needed to treat (NNT) from absolute risk difference
- Know how to estimate the number needed to treat (NNT) from Kaplan Meier survival curves.

João Pedro Ferreira, MD, PhD
Nicolas Girerd, MD, PhD
Centre d'Investigation Clinique 1433 module Plurithématique
CHRU Nancy - Hopitaux de Brabois
Institut Lorrain du Coeur et des Vaisseaux Louis Mathieu
4 rue du Morvan
54500 Vandoeuvre les Nancy

[chelseabradford]

for NNT to harm, is that NNH? when I calculate my ARR is positive which is NNT but the article I'm evaluating it should be NNH. But I can't figure out what I am doing wrong!