The Annals of Bad Research; public health edition


A quite delightful misinterpretation of a piece of public health research over at Salon:

States with lower HPV vaccination rates have higher cervical cancer rates Science confirms what we probably could have guessed

Well, no actually, Science would tell us that HPV vaccination rates would have no effect whatsoever on current rates of cervical cancer.

A new study presented at a conference for the American Association for Cancer Research found that states with the lowest rates of human papillomavirus (HPV) vaccination have the highest rates of cervical cancer and deaths from the disease. This is not surprising.

Well, actually, yes it is.

Low rates of HPV vaccination in Southern states have troubled medical professionals for some time.

Ah, yes, all those religious (and possibly even Republican!) Southerners not vaccinating their daughters.

What the paper itself actually says is that those places which have more preventive medicine have more preventive medicine. Vaccines are preventive medicine: as are things like Pap smears which can find potential cancers that can be treated before they become cancers.

But there's absolutely no causal connection at all between high vaccination rates and the subsequent lower cervical cancer rates. As science would tell us.

For cervical cancer takes 10-20 years to develop. And Gardasil, the first HPV vaccine, has only been on the market since 2006. To a reasonable approximation exactly no cases, yet, of cervical cancer have been prevented by the vaccine. It's also true that the likely age group to present with the cancer is women between 35 and 55. Absolutely none of whom will have had the vaccine as it is not offered to those who have already become sexually active (again, to a realistic level the number of those in that age group who were virgins in 2006 is going to be zero or darn close).

The vaccine itself is a wonderful idea and we thoroughly support everyone who will benefit from it getting it (and that includes men too, on the grounds that the tango does take two). However, let's not make up stories about it all. An 8 year old vaccine for a disease that takes 10-20 years to present will have had, as yet, absolutely no effect on the numbers presenting with that disease.

Some might think this not important but come along now, we've got it on good authority that comment is free but facts are sacred.