... dioxin related points.
While I was in my car a few minutes ago I heard on the radio a classic example of a bad science paper getting big publicity. If you're in science and publish, and think about these things, you will know of lots of examples of bad science getting into peer reviewed journals. But I suddenly realized that if you see science only at a distance, you could end up with overly optimistic viewpoints on the motivations of scientists, on their abilities, and on the accuracy of what gets published and reported.
The obviously bad science report on the news claimed that fracking causes health problems. Because this is politicized science, it's going to get huge airplay. Here's what's being reported:
For the study, the team of researchers took a look at the top 25 specific medical categories for 18,000 hospitalizations in the Bradford, Susquehanna and Wayne Counties. Records revealed data from 2007 to 2011.
The team noticed that people living in the Bradford and Susquehanna areas we more likely to be hospitalized for heart problems by about 27 percent, compared to people living in Wayne County.
From 2007 to 2011, fracking increased significantly in Bradford and Susquehanna. These areas have a fracked well density of over 0.79 wells per square kilometer. In Wayne, however, fracking is not allowed because the county is located near the Delaware River watershed.
"At this point, we suspect that residents are exposed to many toxicants, noise and social stressors due to hydraulic fracturing near their homes, and this may add to the increased number of hospitalizations," said Dr. Reynold Panettieri, Jr., deputy director of the Center of Excellence in Environmental Toxicology at the University of Pennsylvania, and senior author of the study.
http://www.techtimes.com/articles/6...ing-sites-have-more-health-problems-study.htm
In other words, the team looked at "25 top medical categories" and found which ones were positively correlated with fracking. This is what's known as a "fishing expedition". Based on random chance, half of the 25 top medical categories will be positively correlated and half will be negatively correlated. If you look at enough variables, you are almost certain to find some that correlate the way you want it to and to whatever degree you want. You then ignore the categories that *didn't* correlate and only make a big deal out of the one that did. This is a classic example of garbage science. No indication of what chemical, how much exposure, or how the effect occurs biologically.
So I went to the journal and read the crap paper here:
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0131093
From the above website, here's their table of "Risk Ratios". I've circled the cardiology result in red:
The "p-value" shows how significant the result is. Smaller p-values mean more significant (i.e. less likely to arise from chance). You can see that the risk ratios are all extremely close to 1. This is a great example of a situation where the signal, if any, is so small it will be swamped by the noise.
Suppose someone wanted to show that fracking was good for your health. Of course they would massage the data differently and get numbers that were more supportive of their political cause. But even if they used the same data analysis that was done in the above paper, they would have the result that the risk ratio for Orthopedics (which I've also circled in red) is *lower* as a result of fracking. And it's lower by the same amount that cardiology went up. And the p-value is 0.011 which means that the result could happen by chance only 1.1% of the time. But did they put out a press release showing that fracking reduces orthopedic risks? That would be "no", LOL.
Now how much risk are we talking about here? The quoted risk ratios for Cardiology and Orthopedics are 1.0007 and 0.9993. I laugh to see numbers this close to 1. It means that in the highly unlikely event that their paper made sense, the people being admitted for Cardiology would have increased by 7 for every 10,000 cardiology admissions, and the admissions for orthopedics would decrease by 7 for every 10,000 orthopedics admissions. This is in complete contradiction to their headline which claims a 29% increase in admissions, LOL.
Now put on your thinking cap. What kind of changes happen to a county where they start producing a lot of gas? The biggest change I can think of is *not* that there's chemical exposure. That is minuscule and is controlled by the EPA. No, the big changes are the direct result of all those fracking jobs. A lot of people start getting paid a lot of money. They start working hard in the field. The unemployment rate goes down. People who own land get paid for the rights to dig. Roughnecks move into town. The economy picks up and some people probably move permanently into the area.
In fact, here's the official FRED economic data for the unemployment rates in Susquehanna County, (where fracking was done a lot and is used in the above story), and in Wayne County (the county they used for comparison, where there's no fracking):
I've circled the region where fracking decreased the unemployment rate in Susquehanna County. Note that this is in complete contradiction to their paper's claim: "
There were no striking differences among the three counties. The subjects were predominantly Caucasian with few people obtaining higher than a high school diploma. Further, the median income was similar among the counties." The truth is that there *was* a striking difference between (at least two of) the counties. The unemployment rate! And of course, it would follow the amount of fracking. Fracking brings jobs.
In other words, they could have taken exactly the same data and used it to show that employment causes an increase in cardiology risk. Why is that?
Could it be that when people get employed and have healthcare, they're more likely to go to the doctor to find out if there's something wrong with their heart? And of course orthopedics goes down because people without children are a lot more likely to move to a fracking area for the jobs. Same with women's heath issues.
I mean really, these guys are complete idiots. No excuse for this to get past peer review. I expect to see it torn to shreds in the next few days.
