Statistical significance

Recently there was a report that BRAC1 defects that lead to TNBC is less deadly than 'wild-type' TNBC (75% survival after 5 years versus 50%). What!?! In the past, BRAC1 carriers were counselled to have mastectomies and their ovaries removed even before any cancer appeared as they were at such high risk for getting breast cancer.
Summaries of this report did not make any sense to me so I read the original report today.
The study took 77 patients with TNBC. No mention of what stage they were or if their treatment was the same. These patients did not have a family history of breast cancer but they were tested anyway for the genetic defect. 19.5% were found to have it despite the lack of history (it is an autosomal dominant trait: you had to get it from one of your parents and there usually is a string of  prematurely dead realatives in your history). One of the conclusions the study made was that everyone with TNBC should be tested; not just those with a history.
They followed these 77 patients. After 5 years, 50% of the non-BRAC1 patients were dead but 75% of the BRAC1 ladies were alive. So it was concluded that BRAC1 TNBC is less deadly. Fair enough. But 19.5% of 77 is 15. Instead of the expected 7 or 8 deaths, there were only 4. My conclusion would be that their sample was way too small. My secondary conclusion was that not all BRAC1 defects are the same and the fact that it didn't kill their relatives suggest they have a less deadly type. Also the fact that TNBC killed 50% of women just like myself wasn't a day brightner but maybe they didn't have the same treatment as I did or were a higher stage.
I got my hair done today. The chemo curls are becoming laxer and laxer. I still have a little wave left though, my cancer gift.
I bought this very cool jacket at a boutique on the way there. It's pumpkin colored. Hopefully I won't be confused with the great pumpkin.