Professeur Universitaire de Médecine Générale, Lyon 1.
Médecin Généraliste
PhD, philosophie de la medecine
Aucun conflit d'intérêt.
Mar 7 • 5 tweets • 1 min read
@dnunan79 @LGHemkens @RecoveryDoctor @GuyattGH @AnilMakam For any widespread intervention (screening, treatment): yes. Because the risk of false positives has significant consequences. But for other issues, yes, other criteria for evidence must be established.
The important thing is to set them in advance. For example :
@dnunan79 @LGHemkens @RecoveryDoctor @GuyattGH @AnilMakam To determine whether a treatment can be harmful, more than just high-quality RCTs are needed...:
Data triangulation...
But criteria are needed to assert that the treatment can be harmful.
Mar 6 • 8 tweets • 2 min read
@dnunan79 @LGHemkens @RecoveryDoctor @GuyattGH @AnilMakam Finally, here is where our disagreements stem from: I take as a model the regulation of medicines, which sets the rules for marketing authorization: these rules therefore establish criteria for evidence (consensus), which are, incidentally, subject to change.
@dnunan79 @LGHemkens @RecoveryDoctor @GuyattGH @AnilMakam But that is what is required for EBM... otherwise, everything is evidence! jamanetwork.com/journals/jama-…
