Thanks Evo. Like all models, none are correct but some are useful. All models for CVD risk prediction suffer from this issue as you well know. But we still have to make a decision. We can update the model with more risk variables, which is what we do in real life. But we still have to make a decision.
Thanks Evo. Like all models, none are correct but some are useful. All models for CVD risk prediction suffer from this issue as you well know. But we still have to make a decision. We can update the model with more risk variables, which is what we do in real life. But we still have to make a decision.
Thank you very much for your reply, it is a pleasure to read your thoughts. I think it is relevant to use powerful algorithms, not necessarily with causal markers, which we are already having with a very good AUC, maybe as many authors are proposing 30 years, better than 10 yr. I think it is a drama that many people do not medicate because of the noise of the naysayers, but it is also very relevant to overmedicate.Thanks again.
Thanks Evo. Like all models, none are correct but some are useful. All models for CVD risk prediction suffer from this issue as you well know. But we still have to make a decision. We can update the model with more risk variables, which is what we do in real life. But we still have to make a decision.
Thank you very much for your reply, it is a pleasure to read your thoughts. I think it is relevant to use powerful algorithms, not necessarily with causal markers, which we are already having with a very good AUC, maybe as many authors are proposing 30 years, better than 10 yr. I think it is a drama that many people do not medicate because of the noise of the naysayers, but it is also very relevant to overmedicate.Thanks again.