Comments on: An upper age limit for screening mammography? http://www.kevinmd.com/blog/2008/04/upper-age-limit-for-screening.html Tue, 14 Feb 2012 17:18:00 +0000 hourly 1 By: Anonymoushttp://www.kevinmd.com/blog/2008/04/upper-age-limit-for-screening.html#comment-85235 Anonymous Thu, 24 Apr 2008 14:50:00 +0000 http://clients.emmense.com/kevinmd/2008/04/an-upper-age-limit-for-screening-mammography.html#comment-85235 <i>Five-year breast cancer-specific survival rates were 94% for regular mammography users, 88% for irregular users, and 82% for nonusers (P<0.0001). </i><br/>Does anyone of these researchers understand lead-time bias? overdiagnosis bias? length bias? 5-year survival numbers ONLY make sense when you start from the same point of origin. If a train crashes in Miami, I take it in NYC and you take it in Washington, then my survival is longer. Yet you wouldn't say I am better off? <br/><br/>Any screening, regardless of whether or not it saves lives results in improved 5-year survival rates. I bet neuroblastoma screening in Japan - a major fiasco that failed to save a single kid but harmed a few - also increased survival rates by simply increasing the number of diagnosed cases.<br/><br/>How can published researchers continue to ignore lead-time bias? ONLY comparison of numbers of people who died of breast cancer in screened vs not screened group makes sense when evaluating if screening has benefit. Five-year breast cancer-specific survival rates were 94% for regular mammography users, 88% for irregular users, and 82% for nonusers (P<0.0001).
Does anyone of these researchers understand lead-time bias? overdiagnosis bias? length bias? 5-year survival numbers ONLY make sense when you start from the same point of origin. If a train crashes in Miami, I take it in NYC and you take it in Washington, then my survival is longer. Yet you wouldn’t say I am better off?

Any screening, regardless of whether or not it saves lives results in improved 5-year survival rates. I bet neuroblastoma screening in Japan – a major fiasco that failed to save a single kid but harmed a few – also increased survival rates by simply increasing the number of diagnosed cases.

How can published researchers continue to ignore lead-time bias? ONLY comparison of numbers of people who died of breast cancer in screened vs not screened group makes sense when evaluating if screening has benefit.

]]>