Responses
Letter
Effect size in CASTLE-AF trial: the issue of ‘the tiny effect’
Compose a Response to This Article
Other responses
Jump to comment:
- Published on: 14 January 2019
- Published on: 14 January 2019effect sizes
In general there is often not enough attention paid to the difference of clinical and statistical significance. I do not understand, however, how the authors of the letter conclude that the clinical effect in CASTE-AF was "tiny".
The application of Cohen's d to a study with a discrete outcome is rather unusual. A more intuitive approach, if one is unfamilar with hazards, would be to use the relative risk as outcome measure. The risk of the primary outcome in the ablated group is 28.5%, the risk in the comparison group is 44.6%. The relative risk is 0.64. The absolute risk reduction is 16% and the number needed to treat is 6.25 (for three years). Not a tiny effect.Conflict of Interest:
None declared.