Thursday, April 11, 2019

MUCH MORE ON EARLY DETECTION


Cordova, Alaska, Coast Guard detachment

My daughter Kristen once echoed my sense of frustration, writing “Why isn’t there a simple blood test that will tell you whether you have cancer?”  At that time I was volunteering for a big-time research project aimed at developing exactly such a test.  We worked on it for several years, spent $8 million of your tax dollars – and failed.  Turns out the problem is not as simple as it ought to be.

The link below will bring you up to date with progress along this line.  The specific question addressed is: if you have a lump of stuff in your uterus, should you be referred to a specialist for possible surgery?  To answer that question several blood-sample tests have been developed.  Clearly, if CA 125 is elevated they send you off to a specialist clinic.  The other tests described in this article depend on multiple proteins in the blood plus menopausal status; as such they actually are computer algorithms.  One you are familiar with (if you read my blog) is ROMA: the Risk of Malignancy Algorithm.  Several more are described, possibly better than ROMA in some ways.

The bottom line is that your local doc should know about these things and make efficient use of them.  The article below makes it clear that applying these tests to the common asymptomatic woman who wanders in off the street is “inefficient” and is not recommended.  My take is: if it’s cheap and easy, go for it.

If you read this link (big “if”, I’m afraid) you might need to be reminded of some medical stat things:
Sensitivity: the probability that, if the test says you’ve got it you really do
Specificity: The probability that, if the test says you don’t have it, you really don’t.
PPV (Positive Predictive Value):  The number of true positives, divided by the sum of true positives and true negatives.
Bottom line: the bigger these things are, the better

No comments:

Post a Comment