Why Odds Ratios Can Be Tricky Statistics
November 2018
in “
The Journal of Clinical Psychiatry
”
TLDR Odds Ratios can be misleading; it's better to also present risk and Relative Risks for clarity.
The document from November 27, 2018, discussed the use of 5α-reductase inhibitor drugs finasteride and dutasteride in treating benign prostatic hyperplasia (BPH) and their associated sexual adverse effects. A meta-analysis of 17 randomized controlled trials involving 17,494 men showed a 69% increase in the risk of impaired libido and a 55% increase in the risk of erectile dysfunction in men with BPH. However, these risks were not significantly impacted in men with androgenetic alopecia (AGA), possibly due to lower doses used. The document also highlighted the complexities of interpreting Odds Ratios (ORs) and Relative Risks (RRs) in statistical analyses, emphasizing the importance of presenting risk and RRs wherever possible.