Thanks, Walter for informing me about the De Hert article!
Thanks for trying to move attention away from the statistical significance framework, which is not helpful for patients. I agree that NNT is more informative, but NNT must always be related to specific outcomes, not to unspecific “improvement”. And number needed to harm (NNH) must be included in the benefit-harm equation. Each patient has his or her own weighting of every possible outcome, positive as well as negative. Some adverse effects are common, and others are extremely rare. Being exposed to the risk of adverse effects is not the same as experiencing these effects. We need more information about NNTs but perhaps even more about NNHs and how patients value their ratio.
Thanks, Walter for informing me about the De Hert article!
Thanks for trying to move attention away from the statistical significance framework, which is not helpful for patients. I agree that NNT is more informative, but NNT must always be related to specific outcomes, not to unspecific “improvement”. And number needed to harm (NNH) must be included in the benefit-harm equation. Each patient has his or her own weighting of every possible outcome, positive as well as negative. Some adverse effects are common, and others are extremely rare. Being exposed to the risk of adverse effects is not the same as experiencing these effects. We need more information about NNTs but perhaps even more about NNHs and how patients value their ratio.