Help Me Reach 12 on the Manly Scale of Absolute Gender

If you like the patriotic work we're doing, please consider donating a few dollars. We could use it. (if asked for my email, use "")

Thursday, November 01, 2007

Statistics and Rudy GiuLIEani

I see that Rudy is going to continue to use his so-called statistics to justify his position on health care. Just so we're clear - statistics can be used to justify ANYTHING. All you have to do is cherry-pick, ignore confounding factors, etc.

I have been writing on another thread regarding all the misconceptions Rudy's misuse of these numbers is bringing up.

First, the parameters must be the same. Rudy is using a set of figures from Britain that comes from their version of the CDC. The US numbers are from a right-wing think-tank paper that doesn't even claim to use actual numbers to justify the result. Apples to hand grenades anyone?

Secondly, before one can even begin to quantify the "quality of care" you must adjust for the natural variability in the studied population. The US is a melting pot - we have large percentages of all kinds of different racial and ethnic backgrounds. Most other countries in the world are not that way. They have a large group of their "founding" ethnicity, and a small minority of immigrants. This all has to do with basic genetics. (An example is that southeast Asian cultures have a much higher rate of lactose-intolerance than western Europeans.)

Third, one must adjust for attitudes. In the US, we have a societal abhorrence for the very natural part of life that is called death. We collectively feel that we must fight to the last inch before giving in to the inevitable. In other areas of the world, this is not the case. People choose to spend their last time with their families instead of in hospitals hooked up to machines. These attitudes do make a difference in outcomes.

Finally, we must all understand just what mortality figures really are - in Rudy's case, mortality rates for people with prostate cancer. A mortality rate is the number (not the percentage) of deaths per 1000 of the population that are diagnosed with the condition. That last bit can really change things.

In the US, the portion of the population covered by any kind of insurance is routinely screened using the PSA test. Lots of prostate cancer is found - however that in and of itself is misleading on the face of it. Prostate cancer comes in two forms, a virulent, fast-growing and quickly fatal variety that affects about 5% of the population, and a slow-growing type that almost 80% of autopsies reveal when the patient has died from other causes. The slow-growing type responds negatively to aggressive treatment, that is, it can respond like the more virulent form if you attack it with chemo or surgery, etc. Most experts now do a biopsy and then advise "watchful waiting" rather than any aggressive treatment.

Of course, the discovery, treatment and outcome of those without insurance is totally different. PSA tests are not administered in the emergency room. People with virulent prostate cancer wind up in the terminal stage of the disease, in the emergency room when it is too late to do anything except make them comfortable. Those with the slow type are never diagnosed at all.

In Europe, PSA tests are not routinely done. Doctors in many of these countries believe that the use of the "unreliable" PSA test results in too many false positives, and aggressive treatment protocols where it is not warranted. (There is also the issue of underfunded health care systems but that is an argument for another day.)

Here we have three groups of people - as you can see - they are not the same. Rudy is trying to convince us that they are. The first two groups here in the US are actually one group - and the one in Britain is the same as our one group. Nope.

It would take a large volume to explain why we may or may not be better or worse than Country B or Country C because of all the mitigating, aggravating and confounding factors that must be accounted for in order to make statistics like those quoted by Rudy meaningful in the least.

Far more useful for the health care discussion would be the mortality rates of the two different groups here in the US. It would answer the question - is some kind of universal healthcare coverage better than nothing? And that, my friends, is really the question. Not whether we are better or worse than Country B or Country C, but what is actually happening right here, right now.

No comments:

Post a Comment

We'll try dumping haloscan and see how it works.