Words Mean Something: The Prostate Story

The New York Times today has an editorial on the prostate papers in NEJM which we commented upon yesterday. The Times says:

"The studies — one done in the United States, one in Europe — both show that screening had little or no effect in reducing prostate cancer deaths."

That is NOT what the papers said. They said that the protocols used to screen had little or no effect. NOT that "screening had little or no effect". Really folks, words count.

The question the researchers should have asked was:

"What level of PSA yields a positive result regarding the reduction of mortality?"

or even better:

"What level of PSA and what level of PSA velocity yields a positive result regarding the reduction of mortality?"

They did not ask that question. They asked the question:

"Does a PSA test of 4.0 threshold reduce mortality as compared to two sample groups."

Well, as we also said the American sample groups were both "tested" albeit not as frequently, and the European sample groups were for all purposes untested. Thus frankly the level was wrong, which was known since 2003 as in NEJM, in the paper by Punglia et al, which showed that a PSA of 2.3 was required to get reasonable levels! The 4.0 level was outdated for six years. No wonder there was no positive result, in addition to the samples used.

Why worry about stupid reporters and editorial staff writers, well because it may become health policy! And that policy can kill. Consider if we did a test that said for women we screen for palpable breast lesions only larger than 4 cm in diameter. Then we would likely conclude that breast screening is ineffective since those screened and those not screened died at the same rate! Dumb, yes.

This demonstrates two issues:

First, the newspapers do not have the basic competence to read and report the facts. Words mean something and in this case lives hang in the balance.

Second, you may get answers to a question but it may very well be the wrong question. Ten years ago this may have been the right question, but we learned something. So does that mean we just continue a flawed study. I truly hope not.