Giuliani Prostate Statistics - Follow-up Part II

After my "Follow-up Part I" post my liberal friend told me I clearly don't get the point of the Michael Dobbs article. My friend states:

"Take a look at what the survival rate means and how it is measured. It is not the inverse of mortality rate. The survival rate is not what is important. The mortality rate is what is important. The mortality rate is the rate at which people actually die from the disease. The survival rate, as Dobbs carefully points out, is a function of many things, in this case early screening. Since the mortality rates are the same, all early screening does is change the survival rate (as in his example early detection means that, regardless of treatment, you have a better chance of being alive 5 years from now simply because since the disease progresses slowly, not due to medical intervention). This is a function of the early detection. You may say that this is great and you'll take it. Fine, but understand that it DOES NOT affect your chance of dying from the disease - mortality rates are THE SAME. You will know sooner than someone in England that you have prostate cancer, but your chance of dying from it is the same."

I absolutely get the point about morality rates. But my reasoning is that there has to be more people dying of prostate cancer in England that never get diagnosed. You can’t have it both ways. You can’t say that the U.S. survival rates are artificially high due to better screening – and then say that the mortality rate in England is the same as the U.S. when the diagnosis rate, and therefore the mortality rate, clearly does not represent the actual occurrence and consequence of the disease.

How many people die in England because of prostate cancer that was just never diagnosed? Doesn’t it make sense that if there are fewer people diagnosed then the actual number of cases is higher, and the mortality rate is higher as well – and neither are in the numbers?

The diagnosis rate for the U.S. in the Commonwealth Fund report is 132 per 1,000. The mortality rate is 26 per 1,000. The diagnosis rate for England is 49 and the mortality rate is 28. Let’s assume that the U.S. represents the best data for the occurrence of prostate cancer due to superior screening. That means that the rate for prostate cancer in England is not 49 but actually 132 per 1,000.

That means the mortality rate is also higher than what is being reported. If it is never diagnosed, they can’t say you died of it. This is not "CSI: London". They're are not cutting open people that died, but don't have a stab wound, gunshot, etc. If prostate cancer was never diagnosed and the individual dies, it has to be classified as something else: "natural causes, perhaps"?

On a relative basis this means that the actual deaths in England are 75 per 1,000, but 47 of those deaths are not identified as resulting from prostate cancer.

If my friend in his defense of the New York Times and Washington Post op-ed pages can point out a flaw in this I'll pass it along in a subsequent post.