The AP reports that federal officials have "admitted" that there is a "cluster" of blood cancers in a 20 mile stretch of Pennsylvania! Of course, clustering is entirely to be expected in any random distribution and does not indicate any increased health risk; it's just a handy way of scaring people and drumming up readership for your news story.
Indeed, the story is typical of the genre of scary, but scientifically baseless "public health" stories:
The area 80 miles northwest of Philadelphia is home to several Superfund toxic cleanup sites and a power plant fired by waste coal, though the government report stresses not enough data exists to trace the cancer cluster to a specific source.
Given that "Superfund" sites are just about everywhere, and coal-fired plants (waste coal! even scarier!) are just about everywhere also, it's not surprising that you would find such a coincidence. But use the word "admit," and suddenly it looks like a government cover-up has been exposed, and the hysteria can proceed to transform a coincidence into an Industrialized Conspiracy of Death!
The disease in question, by the way, which the story actually never explains, is polycythemia vera, which, being a cancer, is the result of genetic mutations that can be brought on by all sorts of environmental and genetic factors. Nobody knows the exact mechanism. Maybe it really is brought on by the presence of a power plant, or some other unidentified chemical. I doubt it. (Wouldn't PV be far more common than it is if coal smoke were the cause?) But the existence of a cluster of 131 cases in a nation of 350,000,000 people is such a minute data point as to be absolutely irrelevant and distracting.
This world would be a better place if reporters would stop playing epidemiologist.
Update: The Agency for Toxic Substances And Disease Registry notes that "[the] criteria used to diagnose PV have changed with the discovery of the JAK2 mutation. The JAK2 test has only been available to community physicians since 2006. Also, there is incomplete reporting to the registry because currently PV is only reported by hospitals and not all PV patients are hospitalized." And an article in the journal Blood not long ago noted significant differences in the diagnostic testing for PV in different regions of the country, with the physicians in the northeast "utiliz[ing] common tests in the diagnosis of PV" more commonly than physicians elsewhere. It is entirely possible that these numbers are an artifact of diagnostic practice (perhaps doctors in Pennsylvania, eager to use better PV diagnostic tools, only recently obtained those tools?)
Comments policy