The Utility of Information

There's an interesting article in the New York Times today, concerning whether screening for prostate cancer actually does men any good.

Two very large studies have finally come out, one American looking at 77,000 men and one European grouping 182,000 men from several countries, both following their cohorts for at least 10 years. That's just how long it takes to do really solid epidemiological work. You have to identify people and then follow them for years and years to see what happens.

.

It's a long-haul thing, and answers are slow in coming. The temptation is to jump to the "logical" conclusion. Given that cancer is bad (we know that), that the PSA test can indicate prostate cancer (yes, it does), and that early detection means more effective treatment (true for other models of cancer), logically, using a fairly cheap and effective PSA test routinely can save men's lives.

.

Ah, no.

There was no survival benefit for men who were screened vs those who were not.

.

That's because it's turning out prostate cancer is most always very slow growing. It doesn't usually metastasize much. It's a cancer to die with, at a ripe old age, not of.

There are of course exceptions. It can take a wrong turn and become a really nasty, invasive tumor. Some men do die of prostate cancer. But these are exceptions, and the consequences of treating millions of men for something they don't really need to be treated for (until, certainly, they start showing symptoms that indicate a more pro-active approach is required) need to be considered.

Both studies are continuing to follow men, to see if there might be any benefit to PSA screening after 15 or 20 years, and surely other studies are in progress. But for now, it does look like the information gleaned by widespread PSA testing is not useful. Quite the contrary, it may cause unnecessary anguish, painful interventions with occasional lasting disability, and contribute to overloading our health care system.

So we may be in for another shift in 'what people ought to do'. It's hard for people to keep their faith sometimes in doctors and scientists when this kind of shift happens. First there was no test, and some men died of prostate cancer. Then there was a test, and the medical community was shocked to find just how common prostate cancer really was, though about the same number of men died of it. Now we're stepping back and saying, ok, we can identify a lot more cases, but those extra cases are going to be alright anyway. What we still need is a way to find those few men for whom this illness is truly threatening.