I'm interested in conversations about and I want to talk about
Know exactly what you want?
Show search

Notification

Error

A mathematical perspective on survival graphs

User
Posted 10 Jul 2018 at 17:22

Hi all


I have read here and elsewhere (may have misread) that the median for effective HT treatment is 2 years, which sounds very gloomy.


There is a but...


I came across this sight that gives a more positive view on median curves - https://www.cancerguide.org/stats_home.html.


This can be considered in context with https://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/prostate-cancer/survival.


I think it ties in with the fact that most statistics ignore Bayesian elements - my son explained this to me and I got hopelessly lost :)


It also gives some useful PCA links here - https://www.cancerguide.org/prostate.html.


What do other people think?


PP

Edited by member 10 Jul 2018 at 21:29  | Reason: Added URL

Carpe Prostatem

User
Posted 11 Jul 2018 at 18:01
You need to distinguish between data and statistics. The latter is actually the mathematical way of dealing with uncertainty, or figuring out the difference between truth and noise. All clinical trials use the methods of statistics, but that doesn't mean they should be doubted. The data are what they are, and as AC points out, are by their nature historical and therefore subject to improvement via medical science.
Show Most Thanked Posts
User
Posted 11 Jul 2018 at 17:23
Pete, statistics have to be compiled from old data. They tend always to be pessimistic by today's standards. I have a cousin who has been on HT for twenty years - still effective. If he is included in the statistics, which I doubt, as it is an ongoing case, he would skew them somewhat! My advice, doubt what you read if it involves statistics!

AC
User
Posted 11 Jul 2018 at 18:01
You need to distinguish between data and statistics. The latter is actually the mathematical way of dealing with uncertainty, or figuring out the difference between truth and noise. All clinical trials use the methods of statistics, but that doesn't mean they should be doubted. The data are what they are, and as AC points out, are by their nature historical and therefore subject to improvement via medical science.
 
Forum Jump  
©2018 Prostate Cancer UK