To hell with evidence

| 1 Comment

Read this terrific piece of journalism in the New York Times - "Weighing the Costs of a CT Scan's Look Inside the Heart."

It analyzes important questions about the lack of evidence for these tests, the costs, the radiation risks, and the conflicts of interest of many who promote them.

The story includes a quote from a physician who is a heart CT scan promoter - "It's incumbent on the community to dispense with the need for evidence-based medicine."

To hell with evidence. To hell with science.

This attitude always reminds of me of the saying, "It ain't what a man don't know that gets him into trouble. It's what he knows for sure that just ain't so."

1 Comment

What is the appropriate standard to judge medical tests?

1. Efficacy (use of tests improves clinical outcomes)

2. Accuracy (the test accurately measures what they are purported to measure)

Laboratory tests are judged by accuracy and reproducibility and never by their effect upon treatment outcomes. For instance, most tests used today in oncology have comparable "sensitivities" and "specificities."

Pet Scans were not approved because they saved lives in a controlled clinical trial that compared the outcome of patients who received care with or without the benefit of a Pet Scan. They were approved because their performance characteristics (sensitivity/specificity) are reproducible, favorable and provide information to treating physicians.

No test in oncology has ever been shown in prospective randomized clinical trials to improve patient outcomes. The existing standard has always been the "accuracy" of the test. This is true for every single test used in cancer medicine, from estrogen receptors to panels of immunohistochemical stains (IHC) to diagnosing and classifying tumor to Her2/neu and CA-125 to cell culture assays to MRI's, CT Scans, Pet Scans and so on.

So, is there evidence that a CT saves lives? No. No diagnostic/laboratory test ever has.

When you get to the new genetic/molecular tests, even the validation standard that private insurance companies is accepting is "accuracy" and not "efficacy." The essential "proof" is that all they have to do for these tests is that the test has a useful degree of "accuracy," not that the use of the diagnostic test improves clinical outcomes.

However, the validation standard the American Society of Clinical Oncology (ASCO) wants for cell-based profiling tests is "efficacy." The cell-based profiling tests have the same entitlement to be judged by the same validation standard as genetic/molecular profiling and all other diagnostic tests. It must be noted that all types of diagnostic tests are just that, "tests" and not treatment.

That's why ASCO needs to update their tech assessment on cell-based profiling assays. And do it transparently this time!

About this Entry

This page contains a single entry by Gary Schwitzer published on June 29, 2008 5:23 PM.

Would back surgery be approved if it were a drug? was the previous entry in this blog.

The $6 Million Dollar Researcher is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.

Archives

Pages

Wikio - Top of the Blogs - Health

Add to Technorati Favorites