
Startling Disconnect between Reported and Actual Drug-Related Harm in Cancer Treatment

This is today's article from MedPage Today: https://www.medpagetoday.com/cardiology/prevention/84827
An IMPORTANT quote from the article:
The study isn't the first to suggest a disconnect between cardiovascular risk in the real world and in the cancer trials. For example, trastuzumab (Herceptin) was associated with a 2.5% rate of incident heart failure in trials, whereas focused cardiovascular disease evaluation put the number at 20.1%.
Interpretation: Drug approval trial data is purposely minimizing the harmful effects of the drugs - in this case by 800%
In a previous blog I discussed the concept of absolute vs relative statistics. Simply put, relative statistics are NOT meaningful to you. For example, your chance of being struck by lightning is very very small. Relative statistics are how the drug companies report benefit.
So, if 2 people died from a lightning strike last year and this year only 1 died, that is a 50% RELATIVE reduction (1/2*100% = 50%). However, out of 1,000,000 people, the ABSOLUTE reduction was 1/1,000,000*100% = 0.0001%. This is your REAL risk.
Absolute statistics are how the drug companies report harm - ON THE SAME LABEL.
Here is an example label for Lucentis - a bad drug that caused fatal and non-fatal strokes:
