@dominics Fair point, but...... 200million prescribed (that's a lot.) But only 1 thousand 700 deaths that's not a lot. More than that could be hit by a bus. What's that as a percentage? I haven't got a big enough screen to work it out😱 Cheers

@highlander - it is the US and it is their latest epidemic. Good questions. It is important to drill into things. That main takeaway is that there often appears to be a male bias when the data is run retrospectively. Women of childbearing age - 75% or newly diagnosed MS patients - are usually excluded from trials. Those that do have to take powerful meds are really schooled about effective birth control as the risks to a foetus are unknown and unethical to study. 200 million scripts each year for opioids. First, you need to define which drugs - exactly are in that category. Are they all new patient initiations? This is the seriously weak part as many of them, may be repeated scripts and some patients may be taking more than one. 17000 deaths are - I am assuming here as non-science journos doing science can be problematic - directly attributable to opioids. Same question as before. Per prescription, or per user? Are any groups over-represented in the mortality figures? It doesn't make for good headlines to put things into context. It bores the reader ;) I prescribeth thee a bloody great pinch of salt for the times you read figures. Nonetheless, the central premise is interesting. www.theproblemwithdata.com (I wite this)