Tuesday, December 15, 2009

Mammogram Math

Mathematician John Allen Paulos explains:
As we now know, the panel of scientists advised that routine screening for asymptomatic women in their 40s was not warranted and that mammograms for women 50 ...

The exact weight the panel gave to these considerations is unclear, but one factor that was clearly relevant was the problem of frequent false positives when testing for a relatively rare condition. ...

Cognitive biases also make it difficult to see the competing desiderata the panel was charged with balancing. ...

Whatever the role of these biases, the bottom line is that the new recommendations are evidence-based.
Huhh? It is difficult to see the panel's reasoning because the panel did not tell us!

Paulos admits that it is "unclear" how the panel made its decision, and then he just makes up his own reasons. I can do that too.

Let's say 100k women get screened at a cost of $200 each. 2k test positive, and they all get $2k in additional tests. 500 test positive, requiring $20k in treatments. Total expense = $20M + 4M + 10M = $34M.

Now suppose the insurance companies want to save money, and cancel all the screening. The 500 women with cancer eventually come in for an inoperable cancer, and treatment is only $6k each. Total expense = $3M. The insurance company just saved $31M by canceling the screening.

Yes, we can that the authorities were looking at evidence, but were they more interested in saving money or saving lives? Unless they detail the exact weight to all their considerations, we don't know who benefits from these recommendations. It does not help for a mathematician to invent possible reasoning. We need to know what the reasoning really was.

This is the future of health-care in America, I am afraid. Bureaucrats in secret meetings will make decisions to ration health care, and they won't document their rationale.

1 comment:

Anonymous said...

That's pretty much health care today. You have health insurance right? Do they call you and pester you every time you make a claim to insure that someone else isn't supposed to pay for it? Do they make arbitrary and unexplained decisions about why you have to pay for the "test" yourself as it's not covered?