This plot that I made goes back to the original paper by Thompson et al. (2003), that was the basis for slapping the Black Box Warning Label on Dutasteride prescriptions.
pubmed.ncbi.nlm.nih.gov/128...
There are some important points that should be considered:
(1) Thompson only studied Finesteride, not Dutasteride;
(2) The absolute difference in % of men who have high-grade prostate cancer (GS = 8, 9, 10) is very small (only 1%), as shown in the plot: 6% for men on Finesteride and 5% for Placebo. That's within the noise and is not statistically significant.
(3) Thompson et al. speculated in their paper that the large shrinkage of the prostate (~25%) when on Finesteride may cause an error (a "bias") that "affects the evidence".
It is true that the probability of a random biopsy needle hitting a tumor increases as the prostate volume decreases, (assuming that the number of tumors remains the same before and after Finesteride). That factor is called a "detection bias".
pubmed.ncbi.nlm.nih.gov/17848668/
After Thompson et al. was published, doctors and researchers have written papers/comments that the black box warning label on Finesteride and Dutasteride should be removed.
I agree!!
Bob Watson