Host: Anti-Aging Psychologist Dr. Michael Brickey
Guest: Dr. H. Gilbert Welch
Broadcast and podcast on webtalkradio.net. The podcast is also on the links below
I think Dr. Welch is right on—over-diagnosis is one of the biggest problems in medicine and one of the biggest threats to our health and longevity. Further, each year in the US over-diagnosis wastes tens, possibly hundreds of billions of dollars.
To review, Dr. Welch described several ways over-diagnosis happens:
By looking too hard for pathology we may engage in a cascade of further testing and treatment that may do more harm than good. Prostate cancer is the poster child. Early diagnosis can lead to biopsies, radiation treatment, chemotherapy, or surgery that brings the usual treatment risks and can cause impotency and urinary problems. While most men over 60 have some prostate cancer, most do not experience any symptoms and eventually die from some other cause.
For women, annual mammograms starting at age 50 also appear to do more harm than good. In a population of 1,000 to 1,500 with women who receive annual mamograms for ten years starting at age 50, one woman would avoid breast cancer, half a dozen women would be needlessly treated, a third would have at least one false alarm, and more than a hundred would be needlessly biopsied. As treatment becomes more effective, the downside of early diagnosis becomes even more pronounced.
A second cause of over-diagnosis comes from changing the cutoff scores, e.g., blood sugar levels for diabetes, cholesterol levels for heart disease, and bone density scores for osteoporosis. Thus with the decision of a committee, the number of people deemed to have a disease can more than double overnight. While the benefits of treatment are usually obvious for people with extreme scores, there are diminishing returns and greater risks for treatment of people with marginal scores.
With increasingly powerful and pervasive MRIs, CAT scans and other tests, doctors stumble onto findings that prompt more testing and possibly treatment. For example, about half of virtual colonoscopies find abnormalities in the kidney, liver, lungs or other tissue. The vast majority of these abnormalities are benign but it is hard to pass on further testing once they have been identified. In scans of other areas of the body, scans need to consider asymptomatic base rates as high as 50% for disc protrusions and 40% for damaged knee cartilage.
With the exception of using genetics to help determine medications for cancer treatments, at least currently DNA testing seems to either tells us what we already know or gives weak correlations that don’t indicate what we need to do differently.
The bottom line is to be cautious about testing when there are no symptoms and when there are no risk factors such as family history, lifestyle risks, exposure to diseases or toxic chemicals, or health or medication concerns.
Dr. Welch also cautions about survival rate data. A man who is diagnosed with prostate cancer at age 60 and dies and 91 of prostate cancer would have had an impressive 30 year survival rate. If his prostate cancer was diagnosed at age 87 and he died at 91 he would fail the 5 year survival criteria. Testimonials from celebrities such as Bob Dole, General Schwarzkopf, and Arnold Palmer illustrate how early detection and survival rate data can be misleading.