Monday, January 16, 2012

When the cure proves more harmful than the disease

Recently, the Boston Globe published a series of haunting images taken from the evacuation zone surrounding the stricken Fukushima plants. Perhaps what is most striking is the unearthly nature of the abandoned towns, now turned feral, their only inhabitants being abandoned pets and livestock. These are more snapshots from a post-apocalyptic wasteland; it is difficult to imagine these as bustling, populated places less than a year ago.

Several news agencies have taken up the issue of the evacuation zone, referring to it as "poisoned" and "uninhabitable for decades to come." And of course, taking a look at maps such as that put out by the Washington Post, indeed the situation does look quite grim. (For an excellent, authoritative view of the data, the NNSA has continued to publish updated radiation measurements).

The evacuation zone - and the surrounding dire rhetoric - is founded on the idea of maintain a "safe" level of annual radiation exposure, set as 20 milliSieverts (mSv) per year, or 2 rads/year. (Note that the NRC limits for the general public are 1 mSv/year for the general public, and 50 mSv/year for radiation workers). By contrast, an average member of the public would receive on average 4 mSv/year simply from natural and synthetic sources of radiation each year (including x-rays, CT scans, and the like).

Yet without getting into the issue of radiation hormesis (i.e., the theory that low doses of radiation may show an overall benefit), the epidemiological understanding of radiation risk from protracted exposures at levels under 50-100 mSv is extremely murky. (This is not to assert no long-term risk of excess cancers exists for this level, but simply that our currently understanding of this risk is extremely limited). Conversely, long-term doses above 100 mSv have been shown to contribute to an excess risk of cancers.

Two critical questions should arise at this point. First, what risk exactly is the 20 mSv/year limit based upon? For reference, estimates of radiation doses from smokers put their annual doses around 150 mSv/year (due to polonium-210 deposited into smokers' lungs). While clearly no one should endorse smoking as a healthy life choice, it is regardless a risk that individuals freely make each day at that we as a society (reluctantly) accept. In this sense then, why is the government of Japan making a choices of risk for thousands of displaced individuals at thresholds far lower than individuals themselves accept?

Taking the linear-no-threshold model (LNT) at its face, a common value for excess fatal cancers is assumed to be 0.005 fatal cancers / Sv (i.e., 0.00005 fatal cancers / mSv; the value established by NCRP 115). Assuming a dose of 20 mSv/year, this leads to an overall increase in risk of fatal cancer of 0.03 over 30 years - in other words, 1 in 33 individuals can be expected to develop an excess fatal cancer over a period of 30 years. That comes down to 1 fatal cancer per year per 1,000 individuals exposed - a rate approaching statistical noise.

Still sound scary? Let's put this into the context of other common everyday risks:


RiskOrder (death/year)
Meteorite impacts~0.000000001
Radiation risk to U.S. population from nuclear power plant~0.00000001
Radiation from high-altitude flight~0.0000001
Lightning~0.0000001
Radiation death from consumer products~0.0000001
Flood, tornadoes, or earthquakes~0.000001
Death from cholera or whooping cough~0.000001
Train passenger death~0.000001
Natural background radiation~0.00001
Passenger deaths in aircraft accidents~0.00001
Death from leukemia or tuberculosis~0.00001
Brain damage from whooping cough vaccination~0.0001
Overall road accident deaths~0.0001
Death from bronchitis or influenza~0.0001
Death from childbirth0.00024
Lung cancer risk from smoking (20 cigarettes/day)~0.001
Death from all causes at age 55~0.001
Death from cancer, stroke, or heath disease~0.001
Deaths from 20 mSv exposure (Fukushima exposure limit)~0.001

In other words, the absolute limit on exposure established by the Japanese government corresponds to roughly equal that of the mortality risk of an average smoker (assuming conservative projections of low-dose risk). Beyond this, residents are to be permanently exiled from their homes.

Now conceive of a relatively conservative limit of even 50 mSv/year - a little over double the current exposure limit. From the above exposure maps, this would roughly halve the area requiring permanent evacuation ("permanent" in the sense of on the order of 30 years given the half-life of Cs-137, assuming no further decontamination). This would lead to a risk of about 0.0025 cancer deaths per year - 1 in 400. 

While such a certainly not trivial (although based in conservative projections), the question which should be asked is, "Compared to what?" Those evacuated have been deprived entirely of their livehoods - homes, farms, businesses, and ancestral lands. Clearly, they should be made whole by TEPCO, and failing this, the Japanese government. Yet the additional risk of death from cancer should be balanced against the very real physical and emotional costs of a mass involuntary exodus (not even taking into account the economic consequences for both those impacted and the country as a whole). What is the toll in terms of premature death from stress and associated mental health disorders (including alcoholism, depression, etc.) In essence, what is the cost in terms of human life from a forced exile? 

Nor are forced evacuations without cost - recently, Japan's blanket 20 km evacuation order recently came under fire due to its generic nature (ignoring real-time radiation exposure data, showing that contamination was spreading in a northwestern direction due to seasonal winds). As a result, many individuals received a higher dose than that if they'd simply stayed in place (due to evacuating to the northwest). Likewise, these evacuations themselves necessitated leaving homes (areas where radiation exposure is lower) and taking up residence in less sanitary conditions with greater exposure to the outdoors (increasing overall exposure). Thus, evacuation itself is not without risk. Each alternative must be evaluated for its own inherent risks - there is no free lunch.

I am not about to assert that there is zero risk from allowing residents to return - namely because the data regarding prolonged low-dose exposures is too poor to make a definitive assessment of risk. However, what should happen is that residents themselves should be allowed to make this choice from themselves - rather than the government. Perhaps one of the most disturbing images of the Fukushima evacuation zone is the constant vigil of police forces enforcing the no-entry zone; an act done ostensibly for residents' "own good," despite the legions of family pets and livestock abandoned to starvation or a return to ferity.

Within reason, these are risks residents should be allowed to evaluate on their own in an informed manner, rather than having them imposed upon them.