The Big Picture: Keeping Radiation Risks in Perspective

Seurat La Parade

Societal risk reduction is the combination of many small details

One of my favorite movies is Ferris Buehler’s Day Off, and one of my favorite scenes is at the Art Institute of Chicago.  Beginning with a close-up of a bunch of dots of paint, the camera zooms out to show, first, a low-resolution face and, eventually, a painting (“Sunday Afternoon on the Island of La Grande Jatte,” by Georges Seurat, for those of you who always wondered – it still amazes me how quickly you can find this sort of stuff out through Google).  This is not a bad visual analogy for many things, but I’d like to think about it in terms of risk reduction.  Are we so focused on radiation safety and radiological risk reduction that we have lost sight of the larger picture?  In other words – are we concentrating on a dot and thinking it is the entire picture?

What brings this to mind right now are the continuing focus on Fukushima’s radiological impact and the cost to clean it up (and in the process paying scant attention to the thousands of dead and tens of thousands who lost their homes). Also relevant are a few consulting projects I’ve worked on, and some recent articles about the extraordinary measures people are taking to avoid even the slightest dose of radiation.  For example, would you believe that a dentist in the UK was sued for taking panoramic x-rays of some of his patients?  Nobody claimed to have been injured, but some patients were concerned about being put at needless risk. A few years ago the issue of irradiated gemstones – which have trivially small amounts of induced radioactivity – were an item of regulatory concern, forcing many to ask if there is any reason to think that irradiated gemstones should really be competing for our attention with far more potent radiation sources such as  irradiators, nuclear reactors, and orphaned sources.  Then there is the slew of articles discussing radiation dose to patients (including emergency room patients) from medical x-rays, and much more.

Now, I am not proposing that this attention is bad, and I am certainly not proposing that we drop the concept that we should try to keep radiation exposure as low as reasonably achievable (a regulatory concept called “ALARA”).  But I am proposing that we might try to expand the ALARA concept beyond the strictly radiological and try to think about what is reasonable when we consider all of the risks that are faced by an individual or by society.

Let’s take medical x-rays for example.  If I go to the doctor with a headache, I’m not sure that an x-ray is really called for – and almost certainly not a CT or fluoroscopy.  If my doctor sent me for one of these exams, I’d be interested in his explaining his reasoning to me.  On the other hand, when I showed up at the emergency room with severe abdominal pains I had no objections to the radiation dose from the abdominal CT used to determine whether or not I needed surgery. I felt the same way when my young son required x-rays to try to find out why he was having problems breathing.  Whatever risks a medical x-ray procedure may cause are not clinically significant at the time of the procedure; they may become clinically significant years or decades later, but only if the dose was high enough to initiate a cancer.  When faced with a gravely ill or seriously injured patient, shouldn’t the physician’s focus be the immediate well-being of the patient? Proceeding on a course of action without appropriate information is risky.  And these risks are immediate, not longer-term.  The bottom line is that not taking an x-ray also carries with it a risk – the risk of not having needed diagnostic information.  If this latter risk exceeds the risk from the radiation, then the physician should order the radiological procedure.

We can make similar comments about other aspects of radiation exposure.  It may make radiological sense to spend money cleaning up mildly contaminated sites.  But does it make sense for a society in which 1% of the population dies in traffic accidents to spend a lot of money in areas in which there is very little real risk reduction?  How can we justify spending money in areas that have little or no impact on public health, especially when that money is diverted from other areas that can have a real impact?  And when we consider that traffic accidents, childhood malnutrition and disease, shootings, and so forth affect both young and old, whereas cancer is primarily a risk to those of us who are older, it makes even less sense to be focusing excessively on radiological risks.  I am reminded of an oncologist I met in Cambodia who told me that, at that time, they only needed one radiation oncology clinic for the entire country – because so few Cambodians lived long enough to get cancer.  It made me realize that we in the developed world are fortunate that we live long enough to have the luxury of worrying about cancer.  It also made me realize that it makes sense to give everyone the opportunity to live long enough to develop cancer (or heart disease; no reason to play favorites) – by putting our risk reduction efforts into areas that kill young and old alike.  And this may mean reducing our emphasis on reducing radiological risks when the money spent can accomplish far more risk reduction in other areas.

Certainly we have to protect radiation workers and the population as a whole from the potentially adverse effects of radiation – the radiation safety profession and the regulations they follow serve a valuable purpose in doing this.  And we have an obligation to make sure that we do not abuse the environment unduly.  But let’s also remember that, after all of our best efforts, after all of the money that we spend, after all of the time that we devote to risk reduction, the sum of all of the risks in our lives remains 100% – reducing one risk necessarily increases all other risks because we are all going to die of something someday (I don’t want to sound overly somber, but there is no hiding from this fact). That being the case, shouldn’t we focus on the big picture– making sure that the largest number of people have the greatest opportunity to live into old age?  As members of society, shouldn’t this be our goal – to spend our society’s resources to the maximum benefit of everyone?  Yes – radiation safety and radiological dose reduction is a part of this picture.  But let’s remember that it is only one small part of a much larger picture – and that we cannot let our devotion to a single dot of paint overshadow our responsibility to the whole.


Dr Y is a certified health physicist, trained in nuclear power plant design and operations, with experience in nuclear power, environmental science, and planning for radiological and nuclear emergencies. He has 30 years of experience in the areas of nuclear and radiation safety.

Tags: , , ,

9 Responses to “The Big Picture: Keeping Radiation Risks in Perspective”

  1. George S. Stanford December 27, 2011 at 5:19 PM #

    Dr. Y is to be congratulated for his perspective on radiation-related mythology. However, his contribution would have been even more enlightening if he had not implicitly assumed that the LNT model of radiation risk is realistic (it’s not).

    Background: “LNT” stand for “linear, no threshold” — it’s the notion that the health risk per millisievert of radiation dose is constant all the way down to very tiny doses and dose rates. Illustration: Suppose 100 aspirin pills taken at once is the mean lethal dose. Then the LNT model would predict that if each of 10,000 people were to take one aspirin a day for ten days, about 1,000 of them would die of aspirin poisoning. LNT was adopted as the basis for regulations, not because it was established fact, but because it was considered “conservative” and easy to administer.

    While radiation damage does seem to be linear at the cellular level, the factor that LNT ignores is the body’s repair mechanisms, which tend to take care of damaged sells as long as the dose is not overwhelming. If LNT were really operative, the health effects at the locations (e.g. parts of India, Iran, China, South America) where the background radiation is many times the U.S. average would be horrific.

    As a “certified health physicist,” Dr. Y should become aware that not only does LNT for ionizing radiation have no empirical basis, but there is a large body of evidence to the contrary. Just scratching the surface of what’s out there, here are some relevant links:
    - A report from Lawrence Berkeley National Laboratory
    - Testimony before the U.K. parliament
    - J. Cuttler on hormesis

    — George S. Stanford
    Reactor physicist, retired
    Former member of the FAS National Council.

    • Dr. Y December 27, 2011 at 8:38 PM #

      I couldn’t agree more with you about LNT and I would like to refer you to a recent posting about airport security screening in which I indirectly addressed some of the problems of LNT. Having said that, I should also apologize for not directly addressing that question in this posting.

      I suspect that using the term “ALARA” can imply an acceptance of LNT. However, I would suggest that the key term “reasonable” applies regardless of which dose-response model we choose to apply. For example, if there is a threshold dose below which we see no added cancer risk then it is not reasonable to reduce doses below that threshold – therefore, applying “ALARA” would suggest that we not spend extra money to reduce doses to less than that threshold. But even if we do accept LNT “ALARA” does not mean eliminating exposure to all artificial radiation because there comes a point at which our radiation risk-reduction measures are generating harm by taking money from other risk-reduction measures that are more effective. In short, no matter which radiation risk model we choose to follow there comes a point at which it simply makes little sense to continue reducing radiation exposure.

      So – I suspect we are largely in agreement on this point, and thanks for pointing out to me that I should have phrased parts of this better!

      • George S. Stanford December 27, 2011 at 11:15 PM #

        Thanks for the clarification. Yes, we do indeed seem to be in agreement. Regarding ALARA, I notice that, in his statement for the U.K Parliament, Wade Allison suggests that AHARS (as high as reasonably safe) would make much more sense. Amen.

  2. Bob Applebaum December 28, 2011 at 4:38 PM #

    Mr. Stanford is incorrect. LNT is the best dose response model, based on epidemiological data. LNT does not ignore the body’s repair mechanisms, it accounts for them through the use of a Dose & Dose Rate Effectiveness Factor. The results of the LBNL link actually support the use of the DDREF. In fact, LNT is just a simplification of the Linear-Quadratic model. The two are statistically the same in the low dose range.

    His other two links also contain scientific errors, but for brevity I’ll avoid elaboration. I do have an Allison rebuttal here:

    Regarding expenditures, ALARA does contain the word “reasonable”. But we don’t always spend that way whether in radiation protection, national security, or many other programs.

    • Joffan December 28, 2011 at 6:04 PM #

      Bob, your new-found love of DDREF is ill-founded. The whole basis of LNT is that it can infer low-dose effects from the high-dose effects. Introducing DDREF creates a new model, LNT/DDREF, which has no high-dose support at all, and therefore no justification above the null hypothesis of “no effect”. And since normal cancer rates are not even close to zero at zero dose, there is no “leverage” within low-dose epidemiology to demonstrate any effect at all. DDREF is simply a mechanism to avoid admitting to a threshold as experimental observations have demonstrated that LNT is over-predicting health effects.

      Taking it from the top:
      - LNT claims that health effects show a linear response to radiation dose.
      - Low dose observations invalidate this.
      - DDREF is an adjustment to LNT that destroys its theoretical basis.
      - Thresholds account for the observations adequately and should be used.

      • Bob Applebaum December 28, 2011 at 7:04 PM #

        No, that is cartoonish. That’s like saying since apes exist evolutionary biology is false, man couldn’t have evolved from apes.

        No one is saying that health effects are exactly, precisely, without variance proportional to dose. But when you plot the data, the health effects can be modeled with linear quadratic or a linear dose response. There is no difference.

        DDREF accounts for the observations that health effects for a given acute dose, don’t manifest when the dose is lower or fractionated. That’s how science works….the model fits the observations.

        • Dr. Y January 1, 2012 at 12:48 PM #

          Wow – a few days of vacation and I’m impressed with the amount of discussion! Reading this brings a few thoughts to mind – probably enough for an entire posting, but let me try to sort some out….

          I agree that dose rate must make a difference, but I suspect that it is more complex than can be explained by a single factor. To me, the question of LNT vs threshold vs hormesis is like a high-school math bathtub problem. Radiation is inflicting damage on our cells at a certain rate and our biochemistry responds by ramping up our DNA damage repair mechanisms to try to address the damage. If the rate of damage exceeds the rate of damage repair then we accumulate DNA damage that might lead to cancer; otherwise there is likely little or no impact. Similarly, if the amount of damage repaired is less than what was inflicted then there is a possibility of cancer later in life – if more damage is repaired than inflicted we have hormesis and if these are both the same (up to a point) then we have a threshold. The bottom line is that (in my humble opinion!) this all begins with comparing the amount of damage inflicted versus repaired.

          With regards to DDREF, it makes sense that dose rate plays a role – for the same reason that throwing a bunch of pebbles at a person does less damage then dropping a boulder on their head. This is a similar question – is damage accumulating beyond what the body can repair? If – because of the rate at which damage is accumulating or because of the total amount of damage inflicted – DNA damage is accumlating then there is a possibility that this damage might be harmful.

          However- no matter what dose-response model we feel is most accurate – we have to ask ourselves when the added risk becomes important. For example – say a person receives a dose of 1 rem. Using LNT this would give a 0.05% chance of cancer. If there is a threshold of 10 rem then the added risk is 0%, and a hormesis model would call for a slight reduction in cancer risk. But for this level of exposure, is there really a practical difference between 0% and 0.05%? Epidemiology can’t tell a difference, making this more of a philosophical question than a scientific one. And even if the 0.05% added risk is “real” I’m not sure that it makes sense to throw money into trying to reduce this dose when the money can save more lives by putting it into other measures like traffic safety or many public health measures.

          I think I’ll stop here to keep this from expanding to posting-length! And I’ll try to work up a full-blown posting (with links!) in the next few weeks. Thanks, again, for some thoughtful comments.

        • Rod Adams January 2, 2012 at 6:16 AM #

          Bob – I have been struggling with the math that supposedly backs up your assertion of science behind the Linear, No-Threshold (LNT) dose response assumption.

          As I understand it, the LNT assumption is based on the fact that the relationship between dose and risk seems to be linear at the high doses and dose rates that were reconstructed for the victims of Hiroshima and Nagasaki. As we have discussed many times on Atomic Insights, the LNT support community believes that the Life Span Study of that population is the gold standard of radiation dose studies.

          The line that best fits the data at high doses can be extended all the way down to a zero dose, zero risk origin.

          However, later versions of the model add in the DDREF to change the slope of the line at low doses. Can you explain how you can have a line that fits the data points at high doses and then uses a different slope at low doses but still manages to hit the zero, zero origin?

          I cannot, for the life of me, resolve that situation mathematically. The only way to make it work is to add a constant – ie a threshold dose. Maybe my math skills are just too rusty to help me out here. Can you shed any light on the matter of having a line that goes through zero, zero with a fixed slope and then changing that slope without changing any of the data points to still make it through zero, zero?


  1. The big accelerator in the sky | ScienceWonk - August 4, 2012

    [...] of those years according to the dosimeter I bring with me on each flight. And lest you wonder, I’m really not concerned about this radiation – I just don’t consider it a high risk, but since that’s not the point of this posting, [...]

Leave a Reply