If this wasn't so sad, it would be funny.

A piece from the New York Times Opinionator blog on Saturday was a rather self-indulgent rant about x-ray machines used by the TSA in airports. You can find it at: http://opinionator.blogs.nytimes.com/2013/05/25/showdown-at-the-airport-body-scanner/

I found it mildly amusing but the cancer rant got old fast. The consistent usage of "cancer machine" and "radiation chamber" was a bit much, at least for me.

I can grant the author his conceit in avoiding the TSA's look-at-you-nude booths in airports. He is entitled to do so. Most people find the x-ray and milliwave booths obnoxious, including me, and they are probably overkill as a deterrent against high-jacking and terrorism - though that last bit is just my own personal opinion. What I take exception to is the bad science: that x-ray imaging booths will cause gobs of cancer regardless of the epidemiology of cancer risks and regardless of what scientific experts - the kind you can actually cite - say on the subject.

I guess the author thinks that scientific concensus is a pile of crap when compared to his fear of radiation.

He degrades information from experts of radiation and cancer risks - and then turns around at the end of his rant to bemoan the lack of scientific studies on TSA x-ray booths, which he seems to believe will verify his fears. I find this rather odd given that I have had no problem finding peer-reviewed studies on the safety of airport back-scatter x-ray detectors by independent researchers; for example: Mehta, P, and Smith-Bindman, R. (2011), Airport Full-Body Screening - What Is the Risk, Archives Intern Med., 171(12):1112-1115. Now in the academic world of peer reviewed journal, insufficient citation and shallow research to find sources is the dreaded sin of "doesn't know the literature." Basically, this means the author didn't do his homework in investigating his subject; however, journalists are not required to meet the rigorous standards of fact-verification that hard scientists have to endure, and because not enough people challenge what journalists report, they can get away with it. The author of this blog post did actually cite someone though. He cited one other New York Times blog as the sole support for his statements. You can find the cited blog here: http://well.blogs.nytimes.com/2012/08/06/x-ray-scans-at-airports-leave-lingering-worries/

This seems to be a pattern in the blogosphere: one non-expert blog citing another non-expert blog, and it has all the problems that arise when you rely too much on secondary and tertiary sources. I found that the cited blog post was far less annoying than the first but it still made some rather dubious statements. Here's one I liked:

For security reasons, much about how the (x-ray) machines work has been kept secret.

Huh? What? Are you kidding me??? What's secret about back-scattered x-rays? Back-scattered x-rays, also known as Compton scattering, have been around since Compton discovered them in 1922 in St. Louis. He got the Nobel Prize in Physics for it. (ref: http://www.britannica.com/EBchecked/topic/130379/Compton-effect, accessed May 27, 2013).

The technology of back-scattered x-rays is a no-brainer. You pick an emitter and x-ray target combination to get the exact x-ray wavelength you need for the imaging, you design the detector, build a detector-to-computer interface and then write up some software. I could probably design and build one myself, though I would probably have to hire an electrical engineer to get the detector-to-computer interface right and a good machinist to make the mechanical parts.

The authors of these two blog articles both make vague claims regarding secrecy and lack of information on x-ray scanners. I think they might have gotten confused over the potential deployment of mass-transit and van-portable scanners. This concerns a non-airport use of scanners that may be rather invasive toward people just walking around on the street who have no means to decline, protest or even consent to being scanned. The flap might be a real one and a political activist group has filed a freedom-of-information act request to liberate about a thousand pages of paperwork on non-airport use of scanners (see http://epic.org/privacy/body_scanners/Mobile_Body_Scanner_Complaint.pdf). But that's a whole different issue than airport scanners. It has nothing to do with the air-port scanner other than the use of the Compton effect.

There have been some objections to the x-ray booths at airports from some real science and medicine types, like the 2010 publicly-issued letter by four University of California San Francisco (UCSF) researchers. This non-peer reviewed protest over airport x-ray scanners got a lot of press. If these guys had gone through peer-review, they would have spared themselves some embarrassment. You can read the letter for yourself at http://www.npr.org/assets/news/2010/05/17/concern.pdf

The concerns in that letter about safeguards for when an x-ray scanner booth breaks down or malfunctions are legit. The TSA has not been forthright with details on engineering controls and emergency response when breakage happens. But those are engineering concerns and therefore fixable if indeed they are real problems. The letter's science regarding penetration and dose, however, blew it. It looks like four bright profs from UCSF got the basics of received dose vs. penetration wrong. It's an understandable mistake for four medicine-oriented researchers to make. Medical x-rays and material scanning x-rays are different in mechanism and in wavelength. Medical x-rays are directly transmitted and penetrate deeply and through the body. Back-scatter x-rays, also know as Comptom scattering, do not penetrate deeply, but rather they reflect off clothing and skin. The physics is quite different for Compton scattering than for transmitted medical x-rays. The capture of the reflected x-rays are used to make the types of images used in the airport scanners. The radiation dose objections were neatly retorted by the inventor of some of the first back-scatter body scanners during congressional hearings in late 2010. Although it is a bit dense, you can read the rebuttal to the four UCSF profs here: http://www.tek84.com/downloads/Holt-Letter2010-12-2.pdf

The best bit in the cited blog article, at least for me, is as follows:

But some experts are less sanguine, and questions persist about the safety of using X-ray machines on such a large scale. A recent study reported that radiation from the machines can reach organs through the skin.

The author of the cited blog is referring to a Marquette University study (M. E. Hoppe, T. G. Schmidt (2012), Estimation of organ and effective dose due to Compton backscatter security scans, Medical Physics, 39 (6), 3396-3403). The researchers built a computer model and then used it, getting results that a very small amount of x-ray radiation may actually gets through the skin to tissues immediately below the skin. The researchers who published the study did not themselves think this was a big deal since the intensity of the x-rays used in airport scanners is really low, hugely lower than a medical x-ray. Please excuse me for this, but big fat hairy deal! I just love sensationalism in reporting. Don't you? Hey, is that sarcasm I'm smelling?

The cited blog went on further to say:

In another report, researchers estimated that one billion X-ray backscatter scans per year would lead to perhaps 100 radiation-induced cancers in the future.

I confess, it was this statement that convinced me that this was worthy of my blog.

Okay, let's talk about risk. Risk is based on statistics. For real illnesses and injuries, the stats that exist are hard numbers based in the USA on reports sent to the Center for Disease Control (CDC) by doctors. For radiation-based illnesses, the stats that you read about in the newspaper are based on computer models that extrapolate from the very small number of illnesses known to have been caused by ionizing radiation - that includes x-ray caused cancers. The datasets on radiation-caused cancers are not huge. Most of the data is from Hiroshima, Nagasaki, and Chernobyl plus a very small number of nuclear fuel processing exposures and accidental medical irradiations. Because of the sparse data, most of the risk assessments for things like cancers caused by x-rays are based on extrapolations of the sparse data. The problem with extrapolation is that it's based on assumptions of how the data would behave in the gaps where real numbers are missing. If we had all the data we wanted, could we fit a straight line through it or would it look like a curve instead? Those are the sorts of assumptions that go into extrapolations and computer models. And sometimes, the intelligent guesses that modelers make are incorrect. Statistics are only as good as the math behind them.

So back to the study reporting 100 cancers per one billion scans. The study is Brenner, D. J. (2011), Are X-Ray Backscatter Scanners Safe for Airport Passenger Screening? For Most Individuals, Probably Yes, but a Billion Scans per Year Raises Long-Term Public Health Concerns, Radiology, 259, 6-10. That has to be one of the worst journal article titles that I have ever seen but it's a good piece of work that you can read here: http://radiology.rsna.org/content/259/1/6.long (accessed May 27, 2013).

If we go back and look at the cited blog, the author phrased things to imply that the results of the two studies were big and alarming concerns. Let's look at the excerpt as a whole now:

But some experts are less sanguine, and questions persist about the safety of using X-ray machines on such a large scale. A recent study reported that radiation from the machines can reach organs through the skin. In another report, researchers estimated that one billion X-ray backscatter scans per year would lead to perhaps 100 radiation-induced cancers in the future. The European Union has banned body scanners that use radiation; it is against the law in several European countries to X-ray people without a medical reason.

In context, it should be obvious that the author's goal in this paragraph is to imply there is reason to be alarmed, that this cascade of items is evidence of the danger of airport x-ray scanners. Now I'm going to refrain in saying anything about the European Union. I mean, seriously, these are the folks who tried to regulate the shape of vegetables and ban olive oil served in cruets! EU legislators make the current House of Representatives look sane and reasonable! What those people want to do in Europe is not germane to x-ray scanners in airports in the USA. We have already discussed the penetration issue. So what's left is the statistic that the author wants us to believe is a bad thing.

I'm am not going to say anything like 100 extra cancers a year is an okay thing. That's not the issue. The issue here is the perception of risk. Those 100 cancers are not real: they are a statistic generated by extrapolation in an epidemiology model. Cancer will happen to people whether we want it to or not. Numbers like 100 cancers for every 1 billion scans is a risk assessment. It is a reflection of what we think reality will be, but it's not reality itself. There's another thing to consider here, especially when reading stuff like this in the news, and that's the word cancer itself. Cancer is a panic word. It used to be that most folks who got cancer died and the death was unpleasant and painful. Cancer is still a lousy way to die but it is no longer a death sentence. Many cancers now are treatable, survivable and in some cases, preventable. But the very word is still a cause of fright, much in the way that leprosy or polio used to be. The cancer disease group is not the problem here; it's the perception of cancer in American society that's the issue. A journalist need only drop the word into a news report,like the two New York Times blogs here and be assured that the usage will help garner readers. The word cancer in the news is like a giant electrical billboard flashing the fear of the armageddon in a nightmare future that could be our fate. Cancer is a good word to drive readership and readership drives ad revenue.

One hundred cancers per one billion scans: that's an incidence rate, not a death rate. There was no estimate from this study of how many of those cancers would result in death. Now once scanners are installed in every airport, the TSA projects that they will be doing about one billion scans a year, and that translates to 100 cancers per year. That's a projection and it assumes that every scanner uses back-scattered x-rays. We need to translate this into the standard form for disease rate that the CDC uses, cases or deaths per one hundred thousand. The population of the USA is a little over 300 million. Dividing 100 cancers by the population of the country gives us a cancer incidence rate of 0.03 cancer cases per 100,000. Here's another way to say this: in any given year, you have a 0.0003% chance of getting cancer from an airport back-scatter x-ray scanner. You should probably feel really scared now.

Okay, I admit it. I am rather underwhelmed here.

Why is this a trivial number? Easy. Look at real cancer incidence rates for a comparison. It's easy to do so. Check out the National Cancer Institute Fast Stats website at: http://seer.cancer.gov/faststats (accessed on 5-27-2013). It's a cancer statistics tool. I ran the fast stat tool for all incidences of cancer for both genders and all ages. Here's a look at the data for the last several years:


Yep, you read that graph correctly! The current national cancer incidence rate is about 450 per 100,000. That's a number that's based on real numbers of actual reported cancers - it's not from a computer model. So I hope you can forgive my being underwhelmed but the alarming and scary thought of 100 cancers per year from airport x-ray scanners. For comparison, here's some data from the CDC Morbidity and Mortality Datasets. Compared to 100 cancers, more children are murdered every year, between three to four hundred. Between 1000 and 2000 people will be struck by lightning, half of whom will survive. Between 35000 and 40000 will commit suicide. Between 0 to 12 people will be bitten by poisonous snakes, most of whom will live. An estimated 45000 people die every year because they couldn't afford health insurance. One to two people per year in the USA die of rabies every year. Between 30000 and 40000 will die in car accidents; about 10000 of those deaths are due to drunk driving.

People opt out of going through the airport x-ray scanner because of fear of radiation. They should really be scared of travelling in cars.