annals of intern med had article this week reviewing a 5% sample of medicare beneficiaries over 65 years old in different geographical areas and used claims data to assess comorbidities and subsequent survival, creating an updated and more extensive table (see doi:10.7326/0003-4819-159-10-201311190-00005). people with hx of cancer were excluded. the table is reproduced below.
a couple of issues.
–the sample reviewed may not reflect the entire elderly population, since it included fee-for-service medicare only.
–one intention of the study was to provide us guys with a way to assess when to stop health screening, perhaps limiting future mammograms or colonoscopy to those who have a reasonable 5 to 10-year survival. one concern here is that there are no real screening data on people over 70-75 years old (studies not done), so we are left with mathematical modeling, which applies the conclusions for screening younger people (ie that they benefit after 5-10 with early detection of cancer) to the untested older patients. but there are questions about this: eg, is the pathophysiology the same in older vs younger patients (ie, does the breast or colon cancer act the same/follow the same course — i’m not sure here), are the complications of the test the same (here there are data showing that the yield of colonoscopy decreases significantly in those over 75, and the likelihood of colonic perforation increases), are the effects on quality of life the same (elderly bounce back less quickly from an invasive procedure, either because they need a breast bx, or just the colonoscopy itself), do the elderly tolerate the therapy for a detected cancer enough to justify finding one, etc, etc. that being said, i think most of us do continue with screening in older patients with good life expectancy (eg an 80 yo white woman in good health has life expectancy of 11.7 years …. by the way, from this table an 85 yo in good health has the same life expectancy as 66yo with CHF)
–i think it is really useful for us as providers to have a sense of life expectancy anyway, not just for screening, but also in terms of the immediacy of end-of-life planning, ability to talk with the patient about their life expectancy and planning, answer more accurately patient/family questions on prognosis, etc
geoff