<<I don't think you'll ever see a drug data PR without the use of "generally" with respect to side effects>>
Since McBio's post quoted and cited a PR that does exactly that - no generally ("IDX184 was safe and well-tolerated in this study") that seems like an easily disproved statement.