MarkRBol
18.6K posts
@MkRBol
London prison psychiatrist Tweets belong to me and no one else
If the regression model happened to correctly specify the mathematical form (e.g., linear, log-linear, or something entirely different!), includes all relevant variables, and excludes irrelevant ones, it would be SSRI discontinuation! But knowing the correct mathematical form for the regression function *requires* understanding the explanation of why and how various factors cause the outcome *before one can use the regression equation”. In scientific work, predictive (regression) models should come after explanation, not before. So starting with a “regression model” begs the question. It’s circular reasoning. To give you the benefit of the doubt, however, and for the sake of argument, let’s assume we can start with a regression model! But that’s still (!) not particularly relevant in terms of knowing how to help someone who is experiencing withdrawal, which is what a clinician wants to do. It is the explanation of why different interventions are needed that is still necessary for scientific understanding of the given person sitting in front of you, not a statistical rank ordering of causal factors. Again, it is the explanation of why the various factors could be doing in the individual person, even if the regression model happened to apply to the population. Dr. Rege is providing helpful information about understanding why individuals respond differently to interventions that may help. Many of us have successfully treated withdrawal using different interventions (some pharmacological) for different people. In the real world, often you have to figure out which one works. @sanilrege
@JonathanShedler The mechanism is different. Withdrawal is more like a physiological dependence, while discontinuation results from the effect of a neurotransmitter change different from dependence or addiction. It feels bad, but it’s not the same phenomena as withdrawal. 2/end
