Kinsley on mammograms

Michael Kinsley is no longer editor of Slate. He gave up that post

amidst rather a lot of publicity last week, saying that he would continue

at the online magazine as a columnist. That was good news: Kinsley,

when he’s on form, which is most of the time, is one of the most perspicacious

columnists in America. He’s particularly good at using subtle but

decisive logical arguments to cut through standard political rhetoric.

Kinsley’s first column

as full-time writer, however, makes one wonder whether his talent

as a columnist was in some way reliant upon his day job. It’s about

the Great Mammography Debate, and it’s terrible.

In a nutshell: The medical industry in the US is very keen on regular

mammograms for women over 50, and possibly even for women over 40.

But recent research out of Denmark reviewed the available data and

concluded that the benefits of regular mammograms were statistically

zero: women who got tested had the same chance of dying of breast

cancer as women who didn’t.

Since then, a National Cancer Institute advisory panel has concluded

there is insufficient evidence to prove that mammography reduces breast

cancer deaths, although the controversy still rages.

Kinsley’s contribution to the debate is worthless. He starts off

with four paragraphs of sheer irrelevance, about official NIH advice,

dating back to 1997, about when and how often women should have mammograms.

Kinsley then moves on to pretty much his only argument in the whole

piece: one he repeats, but never bothers to argue.

Every day, thousands of American women have mammograms

and some of them discover tumors that would have killed them if the

tumors hadn’t been discovered and removed. Almost everyone knows women

who have survived breast cancer thanks to a mammogram.

Kinsley, of all people, knows that correlation is not causation.

It may be true that almost everyone knows women who have survived

breast cancer; it may even be true that most of those women’s cancers

were discovered by means of a mammogram. It is certainly true that

if a malignant tumor is not discovered and removed, then it is very

likely to kill. But from none of this can we conclude that mammograms

save lives. If the breast cancer would have been discovered in time

by other means, then the mammogram has done no good.

The problem is that advances in treatment have not kept pace with

advances in mammography. A mammogram will show growths maybe 20mm

in diameter – growths easily confused with many other markings

on the film. The problem is that medicine really doesn’t know what

to do when it encounters a growth that small: quite often the best

thing to do is nothing, and see if it grows.

Continues Kinsley:

No one tries to deny that some tumors are caught

and removed that would have been fatal and that some women are alive

as a result. So there’s no serious case that mammograms have no benefit.

The fancy explanations point to alleged disadvantages of mammograms

that might possibly outweigh the benefits.

This is incredibly disingenuous. For one thing, it completely ignores

all of Kinsley’s own throat-clearing at the beginning of the article:

his whole argument from here on in applies as much to 16-year-olds

as it does to 40-year-olds. There is a serious case that mammograms

have no benefit: it has been made, in excruciating statistical detail,

by a pair of Danish scientists. To say that there is none, and to

base that assertion on a statement that has nothing to do with mammography,

is to commit a logical solecism quite atypical for Kinsley.

Kinsley’s column becomes even more unhinged towards the end, however,

after he’s done some really rather bizarre and off-topic chin-scratching

about how much anxiety a woman feels when she’s been diagnosed with

a certain probability of having breast cancer. Are you ready?

Why blame the mammogram for someone’s conscious decision

to have a biopsy or surgery? Why "blame" anyone for deciding

sensibly that a large chance of saving your life is worth a small

chance that it will turn out to have been unnecessary?

It’s obvious that if someone decides to have a biopsy or surgery

on the basis of a mammogram’s results, then the mammogram has caused

the biopsy or surgery. What’s not at all obvious is that said biopsy

or surgery has a "large chance" of being life-saving, and

a "small chance" of being unnecessary. Kinsley simply asserts

here, without citing any evidence at all.

The whole argument in this piece, insofar as there is one, seems

to rest on the idea that the Danish research thinks that the downside

to having a mammogram in some way balances out the upside. But that’s

not what the research says at all: the research says that even if

there were no downside at all to the cost and inconvenience of the

mammogram, not to mention the pain and anxiety of the biopsies and

surgeries which would otherwise not happen, there still wouldn’t

be any upside to having mammograms: people live just as long without

them.

Now, it’s possible that the Danes are wrong about this, although

most of the arguments against them seem to be far too reliant on breast-cancer

deaths falling since mammography was introduced, which is something

that can be explained by advances in treatment as opposed to diagnosis.

But to miss the point entirely, and to start tilting at straw men:

that’s most unlike Mr Kinsley. I think he should go back to editing.