----------------------------------------------------------------------------
----------------------------------------------------------------------------
CHAPTER 24
Chernobyl:
A Crossroad in the Radiation Health Sciences
----------------------------------------------------------------------------
----------------------------------------------------------------------------
This chapter is arranged in 10 parts:
1. Chernobyl's Cancer Consequences -- Integrity of the Data, p.1
2. The Two Keys to Estimating Cancer-Consequences from This Accident, p.3
3. Bottom Line from Our 1986 Estimate of Chernobyl's Cancer Consequences,
p.4
4. Bottom Line from the 1987 Estimate Issued by NRC, p.5
5. Bottom Line from the 1987 Estimate Issued by DOE, p.6
6. Bottom Line from the 1988 Up-Date of DOE's 1987 Estimate, p.7
7. Reason for the Great Disparity, p.8
8. Some Important Comments from the NRC and DOE Reports, p.11
9. The Threshold and Dose-Exclusion: Ultra-Low Cancer Estimates, p.14
10. Beyond Chernobyl: A Much Bigger Agenda in Parts of the Radiation
Community, p.18
Then tables.
----------------------------------------------------------------------------
Introduction:
-------------
This chapter will compare our independent analysis of
Chernobyl's cancer consequences, with three estimates from
influential segments of the radiation community. We will account
for the huge disparity in such estimates. In addition, we shall
provide some new estimates which use the Cancer-Yields developed
in this book, as well as the Cancer-Yields published in 1987 and
1988 by RERF analysts.
In the process, we will suggest how the response by segments of
the radiation community to the Chernobyl accident could have
serious implications -- extending to nuclear issues far beyond
this single accident, and beyond ionizing radiation to other
health issues and to the practice of science itself.
1. Chernobyl's Cancer Consequences -- Integrity of the Data
============================================================
On September 9, 1986, I presented my analysis of (A) the doses
committed for people globally from the Chernobyl accident, and (B)
the estimated cancer consequences from the doses -- namely, a
half-million radiation-induced cancer fatalities. The Chernobyl
analysis was part of a longer paper (mentioned already in Chapter
18, Part 1) which I presented at the Symposium on Low-Level
Radiation at the 192nd National Meeting of the American Chemical
Society -- the ACS.
The analysis was one of the very earliest detailed estimates of
the cancer consequences of the Chernobyl accident, and was widely
reported by the Associated Press, United Press International, and
Reuters. The paper itself (Go86) has been widely distributed in
the USA and abroad by the Committee for Nuclear Responsibility,
inside and outside the radiation community.
As a permanent record, the entire sections dealing with the
Chernobyl accident are reprinted in precisely the form in which
they originally apppeared, as Chapter 36 of this book.
Original versus Revised Dose-Data:
----------------------------------
There is a very special reason for reprinting the 1986
Chernobyl analysis in its exact form as originally presented. The
doses recorded in the 1986 paper are those reported, within the
first four months of the accident, by sources such as the World
Health Organization, the U.S. Environmental Protection Agency,
several separate country reports, and originally by the Soviet
Union itself (citations are in Chapter 36). There may be good
reasons to have more confidence in these original reports than in
the many revisionist efforts.
The Chernobyl accident dismayed the promoters of nuclear power
in virtually every country on the globe. After the accident, there
has been a continuous effort, by governmental and private arms of
the nuclear enterprise, to put the best face on the consequences
of the accident. One way to "improve" the consequences of the
accident would be, of course, to reduce estimates of the public's
radiation exposure from it. For this reason, there is a realistic
basis for skepticism concerning "revised" dose-estimates --
revisions which may continue to appear for years to come.
In short, it is impossible to know which "revisions" of dose
are truly valid, and which are simply window-dressing on behalf of
the nuclear enterprise.
As we shall see in Part 6, the Soviet Union has revised the
Soviet doses downward, which may or may not be correct. Analysts
for the U.S. Department of Energy accept and use the downward
revisions with apparent contentment (Doe88, p.1515-1517).
Elsewhere, however, items like the following news reports make
it exceedingly difficult to have confidence in Soviet candor about
Chernobyl. The numbers are clearly at the mercy of politics.
On March 6, 1989, the WALL STREET JOURNAL (Wsj89, p.A-1)
reported from Ukraine that "Records of" radiation levels [from the
accident] have been deemed so secret that top Soviet scientific
researchers, let alone local residents, can't get access to them"
-- a statement supported by considerable detail in the full
article.
On April 27, 1989, the NEW YORK TIMES (Nyt89b) reported from
Moscow that -- according to IZVESTIA -- the Soviet Minister of
Energy, Anatoly Mayorets, had signed an order strictly curbing
press coverage of nuclear power accidents. According to IZVESTIA,
the new directive designated as classified nearly all reports on
nuclear and conventional power accidents, breakdowns, or
contaminations of any severity. The order prohibits disclosure of
such information in "non-classified documents and in telegraphic
communications, as well as in material intended for publication in
the open press or for export abroad.
And the Health-Data?
--------------------
On July 30, August 9, and August 15, 1989, the ASSOCIATED PRESS
(As89) filed reports from Moscow on the dispute between the
government and scientists in Byelorussia. Because of continuing
exposure from the Chernobyl fallout, scientists were saying that
an" additional 106,000 people currently need evacuation from
Byelorussian villages, whereas the government there was saying
only 11,000 new evacuations would be needed. The Associated Press
cited the official Tass news agency as the source for all its
reports.
On November 13, 1989, TIME magazine devoted a full page to a
report entitled, "The Chernobyl Cover-Up -- Are Soviet Officials
Still Concealing the Truth about the Disaster?" (Time89, p.73).
Among other things, TIME notes that leukemia and other
radiation-related disorders "have allegedly been misreported as
more innocent sounding conditions."
Sadly, all the reports above constitute a reminder that studies
of delayed health effects (including leukemia and other
malignancies) among Chernobyl-exposed Soviet populations could
become grossly distorted by government interference at many
levels.
We commonly hear statements from the radiation community that
observation of the Chernobyl survivors will provide valuable
additional evidence on the magnitude of the cancer effect. (See
for instance, Webs87, p.424; Doe88, p.1517; Ya89, p.160). We ask:
What reason is there for scientists anywhere to trust the INPUT to
studies of the Chernobyl survivors?
One Aspect of the "Crossroad":
------------------------------
There are several aspects to the "crossroad" mentioned in this
chapter's title. One aspect is the choice between credulous
acceptance -- versus diligent exclusion -- of data from any nation
with a world-class record of distorting truth in the service of
state policy, and punishing those who object.
Both the USSR and the People's Republic of China are such
countries. Nonetheless, certain data coming out of both countries
are immediately embraced by parts of the radiation community.
Suppose the data are "doctored" at some step in the system? Is
human health everywhere to be placed at the mercy of possibly
spurious data which can never be verified? I can think of no
protection other than making a presumption of "guilt" instead of
"innocence" until such countries gradually EARN the trust of the
world. Meanwhile, the unfairness to individual, innocent Soviet
and Chinese analysts (who can receive false data without knowing
it) is undeniably another injustice in a long series of injustices
suffered due to such regimes.
A Distasteful Subject
---------------------
Readers may find the subjects of deceit and bias in research
distasteful. So do I.
But they are not imaginary problems anywhere. In this country,
too -- where temptation ought to be less -- standards in health
research have been sinking so fast that, according to an estimate
from the U.S. Public Health Service (PHS), about one out of every
200 principal investigators is involved in some type of scientific
misconduct.
The PHS estimate above is incorporated into a report on the
problem of misconduct in research, issued by the Academic Senate
of the University of California to the entire faculty in November
1989 (Uni89, p.2). The same report also notes:
"A curious fact about known instances of research fraud is that
most of them have taken place in the health sciences" (Uni89,
p.4).
Indeed, in 1986, the American Medical Association decided to
sponsor a "Congress on Peer Review in Biomedical Publication."
According to Drummond Rennie, M.D., Deputy Editor,West of the
JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION (JAMA), and Elizabeth
Knoll, Ph.D., Assistant to JAMA's Editor, one of the reasons for
the meeting was to look into the responsibilities of institutional
authorities and editors in preventing publication of work
involving "fraud and slippery dealing in research" (Renn88).
Additional problems with peer review are well described in Renn86.
The conference took place in May 1989 (Nyt89c).
In 1988, the Association of American Universities felt the need
to release a report entitled "Framework for Institutional Policies
and Procedures to Deal with Fraud in Research," and in August
1989, the U.S. Public Health Service (PHS) began requiring all
institutions which receive PHS grants to develop a process for
investigating allegations of scientific misconduct and fraudulent
research.
The "Downstream" Victims:
-------------------------
The report of the Academic Senate of the University of
California includes an immensely important warning, expressed by
Karl Hittleman, Associate Vice Chancellor for Academic Affairs at
the University of California San Francisco -- the medical center.
Commenting on scientific misconduct-rates like one per 200,
Hittleman said (Uni89, p2):
"It is the view of Congress, and should be the view of the
scientific community, that no amount of fraud is acceptable,
because of the corrosive effects on science and the bad effects on
public trust."
Then the report paraphrased additional comments from Hittleman
as follows: "Regarding science itself, he says, there is a
`multiplier' effect to fraud: Any instance of it can destroy the
worth of related `downstream' research. Worse, fraud can have
potentially disastrous effects on those touched by research -- on
patients involved in medical clinical trials, for example."
How Much Would It Matter?
-------------------------
How much would "slippery dealing" (Rennie's phrase, Renn88) and
spurious data matter in the low-dose radiation health sciences?
How many people downstream would be touched?
In medicine, almost every patient would be affected because of
diagnostic uses of X-rays and radionuclides. In addition, millions
of workers in this country receive occupational exposure to
ionizing radiation.
But the human species as a WHOLE has by far the biggest stake
in an honest evaluation of the risks from ionizing radiation.
Billions of people (many not yet born) will receive exposure
from the Chernobyl accident, and people everywhere could pay the
price if underestimated risk were to become accepted in this
field. Everyone would face nuclear pollution not just from
accidents, but also by INTENTION (see Part 10).
Examination of the Chernobyl accident by this chapter will
illustrate how very small dose-increases for millions and billions
of people produce huge collective dose commitments. This is not
even in dispute, as this chapter shows. The CONSEQUENCES are. The
human race cannot afford serious underestimates of risk in this
field. Readers will understand why, after they have compared
various sets of numbers provided in this chapter.
2. The Two Keys to Estimating Cancer Consequences from Chernobyl
=================================================================
This chapter will compare several estimates of Chernobyl's
cancer consequences, by myself and by others.
No matter who is estimating those consequences, only two kinds
of values are needed in order to make an estimate.
The first value is an estimate of the all-time collective dose
commitment caused by the accident. This value is in the
"person-rad" unit (or some variant, such as person-gray). It is
calculated by multiplying (the average whole-body dose in rads in
each affected country) x (the country's population), and then
summing all these person-rad values to obtain the collective dose
commitment in all countries combined.
The second value is the conversion-factor from dose to
cancer-fatalities. Most analysts make the conversion by using a
Lifetime Fatal Cancer-Yield for whole-body exposure of a mixed-age
population -- though they may call it "risk factor" or other
names. This value is in the units "cancers per 10,000 persons, per
rad (or cGy)," or cancers x 10^-4 persons^-1 rad^-1. Or more
clearly: (cancers / 10,000 person-rads).
Thus in the appropriate equation for radiation-induced
cancer-deaths, both the persons and the rads cancel out, as shown
in the following illustration -- which uses a dose-commitment of
127.4 million person-rads and the Cancer-Yield of 37.313 fatal
cancers per 10,000 person-rads (from Go81).
Sample Calculation:
-------------------
Radiation-induced cancer-deaths =
(dose commitment) x (Lifetime Fatal Cancer-Yield).
Cancers = (127,400,000 person-rads)
x (37.313 cancers / 10,000 person-rads).
Cancers = 475,368 cancers.
One purpose of this chapter is to show whether the grave
disparity, in estimates of Chernobyl's cancer consequences, arises
primarily from differences in dose-estimates, or whether it arises
primarily from differences in Cancer-Yields.
Percent Increase per Rad, and Cancer-Yield:
-------------------------------------------
Analysts (including myself) sometimes express the
radiation-induced cancer-risk in terms of "percent increase per
rad" in the spontaneous cancer death-rate. In Chapter 16, Part 2,
we illustrated the conversion from Lifetime Fatal Cancer-Yield to
"percent increase per rad." Now we will illustrate the reverse. It
requires the estimated fraction of all deaths which are
cancer-deaths in the population under discussion.
Suppose that approximately 17 % of all deaths in a population
are from cancer. Then the lifetime spontaneous cancer death-rate
per 10,000 persons = 1,700. For many purposes, one can omit
adjustment for cancer-deaths which occurred in a mixed-age
population before the radiation exposure. If the percent increase
per rad is, say, 2.0 percent, then (1,700 cancer-deaths x 0.02),
or 34 cancer-deaths is the Lifetime Fatal Cancer-Yield -- namely,
the number of radiation-induced fatal cancers which occur among
10,000 persons over their remaining lifspan after an average
whole-body internal organ-dose of one rad (or rem).
3. Bottom Line from Our September 1986 Estimate of
Chernobyl's Cancer Consequences
===================================================
Collective Dose Commitment:
---------------------------
In Section 7 of our 1986 estimate -- which is now Chapter 36 --
we developed and demonstrated three different methods to estimate
the average per capita dose commitment from the dominant source of
exposure (the radio-cesiums), according to the particular kind of
measurements which a country was supplying during the weeks right
after the accident.
In Chapter 36, the Technical Appendix 2 describes the types of
measurements, country by country.
In that paper, Table 6 (now on page 36-19) provides our
estimate of average dose commitment in millirads per capita,
country by country, along with each country's population. The
countries which are omitted had made no report available for
inspection, and it was not possible to estimate doses by
interpolation from neighboring countries.
Readers who multiply a country's population by its average per
capita dose, and then sum all the person-millirad values, will
find that our estimate of 475,500 fatal cancers (plus 19,500
leukemias) is based on a collective dose commitment of 127.4
billion person-millirads -- or 127.4 million person-rads. The
geographical distribution of our dose estimate is:
EUROPEAN USSR: 56.9 million person-rad.
NON-USSR EUROPE: 65.6 million person-rad.
OTHER, AS INDICATED: 4.9 million person-rad.
As we shall see when we come to dose-estimates by others, it is
important to note that our dose commitments are all-time
commitments (also called "infinite time" commitments). Although
cesium-134 decays with a radioactive half-life of only 2.3 years
(page 36-5), cesium-137 has a radioactive half-life of 30.2 years.
When 151 years (five half-lives) have passed since the accident,
one part in 32 of the cesium-137 released by the accident will
still exist. (See page 36-29 for the estimated time-distribution
of the combined dose from both cesiums.)
Cancer-Yield Conversion Factor:
-------------------------------
Our 1986 analysis uses the Lifetime Fatal Cancer-Yield of
37.313 fatal cancers per 10,000 person-rads, which is the estimate
developed in Go81 from the worldwide epidemiological evidence.
This value appears on page 36-4 in another form -- namely a
Whole-Body Cancer Dose of 268 person-rads per fatal
radiation-induced cancer. The conversion from Cancer Dose to
Cancer Yield is straight-forward:
Number of cancers from 10,000 person-rads
= (1 cancer / 268 person-rads)
x (10,000 person-rads) = 37.313 cancers.
The Cancer-Yield of 37.313 excludes leukemia.
Number of leukemia cases from 10,000 person-rads
= (1 leukemia / 6,500 person-rads)
x (10,000 persons-rads) = 1.54 case.
It might be noted that this value of 1.54 (from Go85) is in
good agreement with RERF's linear value of 1.2 (Pr88, p.460).
There is no science-based reason for applying any reduction-factor
for low and slow exposure, because the leukemia dose-response is
NOT concave-upward when the full database is used (see Chapter 22,
Part 2).
Estimate of Chernobyl-Induced Cancers:
--------------------------------------
As already shown in Part 2 of this chapter, the estimate of
fatal radiation-induced cancers is the product of the dose
commitment times the Cancer-Yield. So:
(127.4 x 10^6 person-rads)
x (37.313 cancers / 10,000 person-rads)
= 475,368 cancers, fatal.
In Table 6 of Chapter 36, this was rounded off to 475,500 fatal
cancers. In addition, approximately one non-fatal cancer is
expected for each fatal cancer produced. The geographical
distribution of the Chernobyl-induced cancers in Go86, Table 6
(Chapter 36, page 36-19) is:
EUROPEAN USSR: 212,150 fatal + 212,150 non-fatal.
NON-USSR EUROPE: 244,786 fatal + 244,786 non-fatal.
OTHER: 18,512 fatal + 18,512 non-fatal.
The combined and rounded number, 951,000 radiation-induced
cancers, does not include additional cancers expected from the
unestimated doses delivered by radionuclides less prominent than
the radio-cesiums, nor does it include thyroid and other cancers
induced by the sizable radio-iodine doses which were received. All
such cancers are additional to the 951,000 cases. The leukemias
are also additional:
Number of leukemias from 127.4 million person-rads
= (1 leukemia / 6,500 person-rads)
x (127.4 million persons-rads) = 19,600.
This was rounded down to 19,500 cases.
So the bottom line from the 1986 estimate is 970,500
malignancies, from the radio-cesium dose.
4. Bottom Line from the 1987 Estimate Issued
by the Nuclear Regulatory Commission.
=========================================
The report named below, and dated January 1987, was issued by
the U.S. Nuclear Regulatory Commission or NRC (our reference
Nrc87):
REPORT ON THE ACCIDENT AT THE CHERNOBYL
NUCLEAR POWER STATION. NUREG-1250.
According to the report's title page, it was prepared by:
Department of Energy
Electric Power Research Institute
Environmental Protection Agency
Federal Emergency Management Agency
Institute of Nuclear Power Operations
Nuclear Regulatory Commission
The report's Chapter 8, "Health and Environmental
Consequences," was prepared by J. Puskin, C. Nelson, D. Janes, and
S. Myers of the Environmental Protection Agency.
Collective Dose Commitment:
---------------------------
The dose commitments in this report appear to be primarily
50-year "lifetime" estimates (Nrc87, p.8-10, 8-13), and are
characterized by the authors as tentative:
EUROPEAN USSR: 50 million person-rem (at p.8-10).
NON-USSR EUROPE: 20 million person-rem (at p.8-14).
How did Nrc87 arrive at these dose estimates? For the European
USSR, whose exposed population is estimated at 75 million people
in the report, the authors accepted the estimates provided by the
Soviets (Ussr86), except that they reduced the Soviet estimate of
dose via ingestion (Nrc87, p.8-10). As for dose in Non-USSR
Europe, the authors say that any estimate "must be regarded as
highly tentative" and perhaps good "within about a factor of 2"
(Nrc87, p.8-13). They make their estimate by excluding Spain,
Portugal, England, Ireland, Denmark, and most of France, and then
estimating that the remaining population of about 350 million
received an average individual dose of about 60 millirems "spread
over a period of years" (Nrc87, p.8-14).
Cancer-Yield Conversion Factor:
-------------------------------
The authors used a Cancer-Yield of 2 fatal cancers per 10,000
person-rads.
They state: "For illustrative purposes in this chapter, the
staff used a risk factor of 2x10^-4 fatal cancers per rad of
(low-LET) radiation to the whole body, corresponding approximately
to the linear-quadratic, relative risk model described in the
National Academy of Sciences `BEIR III' report (NAS, 1980). With
minor modifications, this model has recently been adopted by two
panels of experts as providing a reasonable central estimate of
the risk from low-level radiation" (Nrc87, p.8-6). The two panels
of experts cited are our references Nrc85 and Nih85.
Elsewhere (Nrc87, p.8-10), the authors also state that one
should expect one non-fatal cancer for each fatal cancer induced
by radiation.
Estimate of Chernobyl-Induced Cancers:
--------------------------------------
This is the product of the dose commitment times the
Cancer-Yield. So:
(70 x 10^6 person-rads)
x (2 cancers / 10,000 person-rads)
= 14,000 cancers, fatal. Plus 14,000 non-fatal.
The geographical distribution in the estimate is (Nrc87, pages
8-10, 8-14):
EUROPEAN USSR: 10,000 fatal + 10,000 non-fatal.
NON-USSR EUROPE: 4,000 fatal + 4,000 non-fatal.
5. Bottom Line from the 1987 Estimate Issued
by the Department of Energy
===============================
The report named below, and dated June 1987, was issued by the
U.S. Department of Energy or DOE (our reference Doe87):
HEALTH AND ENVIRONMENTAL CONSEQUENCES OF THE CHERNOBYL NUCLEAR
POWER ACCIDENT. DOE/ER-0332.
Report to the U.S. Department of Energy
Office of Energy Research
Office of Health & Environmental Research From the
Interlaboratory Task Group on Health and
Environmental Aspects of the Soviet Nuclear Accident.
Prepared by
The Committee on the Assessment of Health
Consequences in Exposed Populations.
Authors:
--------------
Marvin Goldman (Chairman), University of California, Davis.
Robert J. Catlin, Electric Power Research Inst.
Lynn Anspaugh, Livermore National Laboratory.
Co-Authors:
--------------
Richard G. Cuddihy, Lovelace Inhalation Toxicology Research
Institute.
William E. Davis, Pacific Northwest Laboratory.
Jacob I. Fabrikant, Lawrence Berkeley Laboratory.
Andrew P. Hull, Brookhaven National Laboratory.
Rolf Lange, Livermore National Laboratory.
David Robertson, Pacific Northwest Laboratory.
Robert Schlenker, Argonne National Laboratory.
Edward Warman, Stone & Webster Engineering.
Collective Dose Commitment:
---------------------------
Although the authors of Doe87 demonstrate the difference
between 50-year and infinite-time dose commitments (for instance,
at their pages 5.33 and 5.35), they choose to use the lower dose
commitment in making their cancer estimates. We will evaluate the
impact of this preference. Their Table 5.16 provides the "50-year
collective dose commitments" which they use. The same estimates
are called "lifetime collective doses" in their Table J.4.
EUROPEAN USSR: 47 million person-rad.
ASIAN USSR: 11 million person-rad.
NON-USSR EUROPE: 58 million person-rad.
NON-USSR ASIA: 2.7 million person-rad.
UNITED STATES: 0.11 million person-rad.
CANADA: 0.0094 million person-rad.
SUM: 118.82 million person-rad.
Doe87 rounds this off to 120 million person-rad.
How did Doe87 arrive at these dose estimates, from what it
calls (at p.vii) the Chernobyl reactor's "violent disassembly"?
As shown in Doe87 Table 7.1, the 50-year dose commitment for
the European USSR is the Soviet's own estimate, including the May
1987 report by the Soviet Ministry of Health (Ussr87a). Doe87
supplies its own estimate for Asian USSR, and calls it "a very
rough estimate" (Doe87, p.5.60). For non-USSR Europe, the
estimates in Doe87 are derived from analytical methods and data
quite similar to our own in Chapter 36 -- except for Doe87's
effort to stop exposure at 50 years.
Beyond the fiftieth year, some twenty percent of the all-time
dose commitment is yet to come -- an estimate from our Chapter 36,
page 29, with which DOE agrees (Doe88, p.1514). Therefore, Doe87's
all-time collective dose commitment would be
(118.82 million person-rad)
= (0.8) x (All-Time Dose Commitment).
148.6 million person-rad = All-Time Dose Commitment.
Cancer-Yield Conversion Factor:
-------------------------------
Doe87 is explicit at page 7.6 about using the risk model
suggested by the Nuclear Regulatory Commission in its
NUREG/CR-4214 Report (our Reference Nrc85). On page J.3, the
authors describe the Nrc85 model as a "composite"
absolute-relative risk model with linear and linear-quadratic
dose-responses, which work out to a "risk coefficient (fatal
cancers / rad) of 2.3 x 10^-4," for "long-term carcinogenesis.
In other words, Doe87 uses a Lifetime Fatal Cancer-Yield of 2.3
fatal cancers per 10,000 person-rads.
Estimate of Chernobyl-Induced Cancers:
--------------------------------------
This is the product of the dose commitment times the
Cancer-Yield. So:
(120 million person-rads)
x (2.3 cancers / 10,000 person-rads)
= 27,600 cancers, fatal. Doe87 rounds off to
28,000.
In the key table, Table 7.11 (Doe87, Chapter 7, page 7.22), the
authors do not mention any non-fatal cancers. The geographical
distribution of the 28,000 "estimated possible radiation-induced
fatal cancers" in that table is:
EUROPEAN USSR: 11,410 fatal cancers.
ASIAN USSR: 2,500 fatal cancers
NON-USSR EUROPE: 13,000 fatal cancers.
NON-USSR ASIA: 620 fatal cancers.
USA + CANADA: 27 fatal cancers.
6. Bottom Line from the 1988 Up-Date of DOE's 1987 Estimate
============================================================
The article named below was published by the journal SCIENCE in
its Volume 242 (December 16, 1988), pages 1513-1519:
"The Global Impact of the Chernobyl Reactor Accident," by Lynn
R. Anspaugh, Robert J. Catlin, and Marvin Goldman. This is our
reference Ansp88.
However, since this article is basically an abbreviation of
Doe87, we shall refer to it as Doe88 in this chapter. The article
itself states the following in its Note 2, in which their
reference (3) is DOE/ER-0332 or Doe87:
"This article is based on work published by the authors and
others for the Department of Energy (3); this reference can be
consulted for methodological details not reported here. The
present article contains several updates to (3); a major one is a
revision of the collective dose commitment reported for the Soviet
Union."
Collective Dose Commitment:
---------------------------
The "collective 50-year total-body dose" given in Table 3 of
this article is 93 million person-rad (we converted person-Gy to
person-rad). The geographical distribution is given below. As
indicated above (and noted already in Part 1 of this chapter),
these authors accept the Soviets' downward estimate of the dose
commitment in European USSR. Anspaugh et al cite Ussr87b and
Ussr88 in this article. The figure in Doe87 was 47 million
person-rad; it goes down to 32 million person-rad in Doe88.
EUROPEAN USSR: 32 million person-rad.
ASIAN USSR: 0.69 million person-rad.
NON-USSR EUROPE: 58 million person-rad.
NON-USSR ASIA: 2.71 million person-rad.
UNITED STATES: 0.11 million person-rad.
CANADA: 0.0094 million person-rad.
SUM: 93.52 million person-rad.
Anspaugh and co-workers attempt to justify using a 50-year dose
commitment as follows (Doe88, p.1514): "We used a time period of
50 years, a standard interval over which to calculate the doses
for lifetime cancer risks. Exposures over the first-year period
and over infinite time were also derived. As an approximation, the
first-year exposure is 10 % of the 50-year exposure, and the
50-year exposure is more than 75 % of the exposure over infinite
time."
As noted in Part 5 above, it is about 80 % .
This practice of throwing away the dose commitment beyond fifty
years is simply an arbitrary way of reducing the cancer
expectation. Anspaugh et al refer to this as "standard," but one
wonders whose standard this might be, and why it is used.
If we were dealing with one set of persons and there were no
"new entries" to the exposed population, it might be a more
reasonable practice, since even the youngest persons in 1986 would
not be very radio-sensitive after age 50. But this is clearly NOT
the situation.
In this situation, we are dealing with mixed-age populations
from which some are departing by death, and into which others are
entering by birth, every year following the accident. New young
people are always being added to the group exposed by the
Chernobyl accident, in great contrast to the A-bomb study, where
no new persons are added to the exposed group over time.
When about 20 % of the radio-cesium dose will occur beyond the
year 2036, it is a mistake to treat that dose as if it did not
exist. Therefore, we shall convert Doe88's 50-year total-body dose
commitment into an all-time dose commitment:
(93 million person-rad)
= (0.8) x (All-Time Dose Commitment).
116 million person-rad = All-Time Dose Commitment.
Cancer-Yield Conversion Factor:
-------------------------------
Although Anspaugh and co-workers are content to incorporate,
into their up-date, new information provided in 1987 and 1988 by
Soviet officials about the dose, they are silent about the new
estimates of cancer-risk provided in 1987 and 1988 by two sets of
RERF analysts. Both RERF reports (Pr87b and Shi88) cast very
serious doubt upon the Cancer-Yield used in Doe87. Both reports
mean that the value used in Doe87 needs to be a great deal HIGHER.
We shall return to this in Part 7.
While Doe87 was explicit about using a Cancer- Yield from
NUREG/CR-4214 of 2.3 fatal cancers x 10^-4 person-rad, Doe88 is
never explicit. Although Doe88 lists "radiogenic risk factors" and
"reduction factors" from NUREG/CR-4214 (Nrc85), Doe88 never states
that the model works out to a Cancer-Yield of 2.3 cancers. The
statement is absent for good reason.
The operative Cancer-Yield used by Anspaugh and co-workers in
Doe88 turns out to be LOWER than 2.3 cancers per 10,000
person-rads. It is 1.87, as we shall show in a moment. The change
is just there, and it represents a decrement of about 20 percent
in the estimate of Chernobyl-induced cancers.
Estimate of Chernobyl-Induced Cancers:
--------------------------------------
This number is given as 17,400 fatal cancers in Tables 3 and 5
of Doe88.
Since the same tables confirm that this number arises from a
dose commitment of 93 million person-rad, the rate is 17,400 fatal
cancers per 93,000,000 person-rad, or 0.000187 fatal cancer per
person-rad. Multiplying by 10,000 to obtain the rate per 10,000
person-rad, we find that the operative Cancer-Yield in Doe88 is
1.87 fatal cancer per 10,000 person-rad.
The distribution of the 17,400 Chernobyl-induced fatal cancers
is stated in Doe88, Table 5:
USSR: 6,500 fatal cancers.
NON-USSR EUROPE: 10,400 fatal cancers.
NON-USSR ASIA: 500 fatal cancers.
USA + CANADA: 20 fatal cancers
7. Reason for the Great Disparity
==================================
Part 7 focuses on two tables. Table 24-A facilitates comparison
of our 1986 estimate with the three estimates already discussed,
and thereby makes the source of the disparities self-evident.
Table 24-B compares our 1986 estimate with additional estimates
based on (A) the new data in this book, and (B) the recent
findings by RERF analysts.
Obviously, we regard the range of estimates in Table 24-B as
the scientifically reasonable range. However, this book does not
ask readers to accept our OPINION. Previous sections of this book
have presented the scientific input which caused us to develop
this opinion. And those chapters did not evade the inherent
uncertainties.
Now it is up to the readers to make their own judgments about
which range of estimates is the more likely to be correct, on a
strictly scientific basis.
The Message of Table 24-A:
--------------------------
People unfamiliar with this field, and unfamiliar with the
details of the various reports, have expressed surprise to me that
the consequences of the Chernobyl accident can be so differently
estimated. And surprise is natural. A range of 14,000 to 475,500
is startling.
Because the low estimates were published subsequent to my own
estimate, it is widely assumed that the radiation dose must have
truly been far smaller than initially estimated, and that this is
the reason for the markedly lower estimates by the radiation
community.
Nothing could be farther from the truth. Table 24-A makes it
almost self-evident that the massive difference in cancer
estimates has practically nothing to do with the issue of
estimated dose commitment from the Chernobyl accident. Indeed, the
Gofman and DOE estimates of collective dose in Column B are
remarkably close. The NRC estimate of dose deserves no attention
at all, in view of its superficial nature (see Part 4). Even Doe87
(at page J.6) heavily criticizes the dose estimate in Nrc87. As
for Gofman-DOE differences in dose estimates for some specific
countries, these differences are of no consequence here, because
the bottom lines in Column D come from the aggregate dose
estimates.
No Mystery:
-----------
There is no mystery about what causes the difference in the
estimates of Chernobyl-induced cancers. The disparity arises
overwhelmingly from Column C -- the Cancer-Yield, or conversion
factor from dose to cancers.
The radiation community uses some Cancer-Yields even lower than
the range of 1.87 to 2.3 shown in Table 24-A. For instance, Doe87
(at page 7.17) reports that UNSCEAR's 1977 value of 1.0 was used
by the U.K. Central Electricity Generating Board to evaluate
Chernobyl-induced cancers.
And Doe87 itself claims (mistakenly) that 1.0 is approximately
the "lifetime fatal cancer risk" produced by the A-Bomb Study in
the T65DR dosimetry (Doe87, p.7.4). For low-dose exposure, the
Doe87 authors are wrong about this by at least l3-fold, as proven
in our Chapters 13 and 14.
The value of 1.0 as a Lifetime Fatal Cancer-Yield was promoted
also by the 1986 president of the American Nuclear Society (see
Chapter 34, Bertram Wolfe), and 1.0 is called the "official"
estimate by BEIR-3 member Edward Webster -- who seems to regard
1.0 as too high (Webs87).
After Chernobyl, we heard the value of 1.0 used too many times
to count.
The NRC and DOE, however, must have been obliged to make use of
the higher estimate issued earlier by the NRC itself, in
NUREG/CR-4214 (Nrc85), some aspects of which are discussed in
Chapter 22, Part 3. The Nrc87, Doe87, and Doe88 reports all claim
that they base the values of 1.87 to 2.3 in Column C upon Nrc85.
Correcting One of the Errors:
-----------------------------
In Chapter 22, Part 3, readers have seen for themselves ONE of
the obvious errors in this Fatal Lifetime Cancer-Yield from Nrc85.
Except for breast-cancer and thyroid cancer (which is rarely
fatal), the Nrc85 risk-value rests on replacement of the
real-world human epidemiological evidence by the preferred
radiobiological hypothesis that dose-response is concave-upward.
The A-Bomb Study has been invalidating this hypothesis for many
years, and Chapter 22 shows that the radiation committees were
aware of this in 1980 already.
Nonetheless, the Nrc85 risk-model rejects both the supra-linear
and linear dose-responses, and erroneously incorporates DREFS for
low and slow exposure. This is no small matter, as we shall see.
As shown by Doe88 (Table 2, p.1515), this Nrc85 model
incorporates a DREF of 0.3 for low and slow exposure -- which
characterizes almost the entire dose commitment from Chernobyl.
This means Doe87 and Doe88 are using a cancer risk-estimate 0.3
times the LINEAR estimate. In other words, just correcting for
this one error would make the Cancer-Yields about 3-fold higher.
So 2.3 would become 6.9, and 1.87 would become 5.6 fatal cancers
per 10,000 person-rad.
Result -- 84,000 Fatal Cancers:
-------------------------------
The corresponding linear estimates of Chernobyl-induced fatal
cancers would also rise by a factor of about three. For instance,
the estimate of 28,000 would become about 84,000 Chernobyl-induced
fatal cancers.
As we said, this is no small matter. Nor is it any small matter
to reject real-world human evidence on dose-response shape, in
favor of a preferred but hypothetical shape.
As our "Crossroad" title suggests, Chernobyl demands evaluation
in various circles of the radiation health sciences. If the
evaluations use unrealistically low Cancer-Yields like 1.0 or 2.0
-- completely at variance with the existing human evidence -- it
is no surprise to me if the nuclear enterprise has credibility
problems (see Part 8, "Chickens Come Home to Roost").
The Message of Table 24-B:
--------------------------
Other analytical efforts at this time are showing conversion
values (of dose to cancers) of 11, 12, 16, 25, 31, 37 per 10,000
person-rad. Table 24-B presents all of them, not just the highest
or the lowest.
The message from Table 24-B is that, when estimates of
Cancer-Yield are scientifically reasonable, they place estimates
of Chernobyl-induced fatal cancers in the range between 140,000
and 475,000, plus an equal number of non-fatal cancers.
The two RERF entries in Table 24-B need some discussion here.
First, their inclusion should not be interpreted as approval of
the non-constant-cohort, non-dual- dosimetry approach currently
used by RERF. The RERF entries are characterized as "realistic" in
Table 24-B because they are tied to real-world epidemiological
observations -- unlike the NRC and BEIR-3 models which are tied to
a preferred but invalidated presumption that human dose-response
would be concave-upward (see Chapter 22).
Second, as far as we know, RERF analysts have made no public
estimates of Chernobyl's cancer consequences. However, we (and
others) are entitled to use RERF's Cancer-Yields in order to
estimate those consequences, just as everyone else has been using
Cancer-Yields from NRC, BEIR, UNSCEAR, and ICRP for the same
purpose.
Readers are reminded, of course, that RERF Cancer-Yields from
the A-Bomb Study are not directly comparable with our own. We
enumerated several of the reasons in Chapter 14, Part 2.
Our own Cancer-Yields from the A-Bomb Study are explicitly
based on low-dose exposure. In Table 14-C, the estimates are based
on linear interpolation between 11 cSv and less than one cSv
(rem). (In the supplemental DS86 analysis, 11 cSv becomes 15 cSv.)
Our other Cancer-Yields from the A-Bomb Study are based on the
best-fit curve (supra-linear), with linear interpolation between 5
cSv and zero dose.
Now let us consider the two RERF estimates in Table 24-B.
Shimizu + Kato + Schull (Shi88):
--------------------------------
In TR-5-88, page 53, Table 19, Shimizu, Kato, and Schull
explicitly confine their estimates of Lifetime Fatal Cancer-Yield
to acute exposure of 10 rems (cSv). We have already commented on
this in Chapter 14, Part 5, "Venturing below 10 Rems." If they
cannot use their curve BELOW ten rems, we wonder why they can use
it anywhere at all. In the region ABOVE ten rems, the
small-numbers problem makes it increasingly unreliable.
The only science-based reason we can imagine, for not
interpolating along their curve below ten rems, would be positive,
credible evidence that the dose-response changes in this little
dose-segment, or that there is a threshold dose below which no
carcinogenesis occurs. Shimizu, Kato, and Schull neither provide
such evidence nor suggest that they believe any exists.
We have already stated (Chapter 14, Part 5) that, in the
absence of contrary evidence or logic, we consider it highly
reasonable and perhaps obligatory for analysts to presume that the
dose-response which derives from the dose-range as a whole ALSO
characterizes the little segment between 10 rems and zero dose.
Moreover, in Chapters 18 through 23, we showed by any
reasonable standard of proof that there is no safe dose or
dose-rate, and no basis for invoking DREFS for low and slow
exposures.
Therefore, it is perfectly appropriate to use the Cancer-Yield
from Shi88 to estimate Chernobyl-induced cancers in Table 24-B.
However, there is not just ONE Cancer-Yield listed in Shi88,
Table 19. Cancer-Yields are given separately for males and
females. For each sex, these analysts present values from their
best-fit linear analysis, with all eight Dose-Groups included and
also with the high-dose groups thrown out. And then they do it all
again, for their best-fit linear-quadratic analysis, with all the
Dose-Groups included and with the high-dose groups thrown out.
All their values are derived from the DS86 sub-cohort,
1956-1985, with an RBE of 10 for neutrons.
The abundance of Cancer-Yields in Shi88 is not surprising, in
view of the authors' finding (Shi88, pp.50-51) that their data fit
linearity and supra-linearity equally well -- when they include
all the evidence, as we do. In other words, the quadratic term was
NEGATIVE in their LQ analysis when they used ALL the evidence. In
order to obtain a POSITIVE quadratic term in their LQ analysis,
they threw away the high Dose-Groups. We have already criticized
this practice in Chapter 14, Part 2.
The value of 12.4 for Cancer-Yield which appears in Table 24-B
is the average of males and females from the LQ analysis, all
Dose-Groups included.
Preston + Pierce (Pr88):
------------------------
Unlike Shimizu, Kato, and Schull above, Preston and Pierce do
not confine their estimated Cancer-Yields to an acute dose of 10
rems. In Pr88, at page 458, it is described as a linear value per
10,000 persons per 10 mSv (one rem).
As Preston and Pierce did in the unabbreviated version of
TR-9-87, they show the effect of using the reduction-factors
(DREFS) suggested by others, but Preston and Pierce avoid any
direct endorsement of their use.
It would be perplexing if they did endorse DREFS, because the
presumption of DREFS is the concave-upward dose-response, and
these authors do NOT find dose-response to be concave-upward. Both
Preston and Pierce are co-authors of Shi87 (TR-12-87). This RERF
Technical Report finds that linearity and supra-linearity (the LQ
model with a NEGATIVE Q-term) fit the data equally well (Shi87,
pp.28-29). The authors comment:
"For those sites other than leukemia and colon, the fitted
curve associated with the LQ model is invariably concave
downwards, not upwards ..." (Shi87, p.29), and "... since the
curvature is invariably downwards when a curvilinear model gives
an acceptable fit, this would imply a higher risk at low doses
than that which obtains under a linear model" (Shi87, p.30).
It is clearly appropriate to use the Preston and Pierce
Cancer-Yield in our Table 24-B, without a reduction-factor, to
estimate Chernobyl-induced fatal cancers.
The value of 11 for Cancer-Yield which appears in our Column C
is their linear result for the DS86 sub-cohort, 1950-1985, with
Dose-Group 8 omitted and with an RBE of 10 assumed for neutrons.
It is found in Pr88 at page 458.
8. Some Important Comments from the NRC and DOE Reports
========================================================
It is self-evident that private and governmental segments of
the nuclear enterprise, worldwide, have an interest in helping the
public to perceive the Chernobyl accident as a non-disaster -- as
the accident which killed 31 people from acute radiation sickness.
(Robert Alexander of the NRC is particularly candid about the
importance of shaping perception, as we will see in Part 9 of this
chapter.)
In Part 8, below, we will illustrate some of the help offered
by the Nrc87, Doe87, and Doe88 reports with respect to perception.
How is perception of the Chernobyl accident related to a
"crossroad in radiation health sciences" ? The answer will become
clear in Parts 9 and 10, and most of our own comments are deferred
until then. Right here, we will only point out that public
perception of the Chernobyl accident might be quite different if
the geographical distribution of the radio-cesium fallout had been
more concentrated.
Distribution of the fallout was a matter of chance. For
instance, if the rain and wind conditions had been different
during the accident, the same amount of fallout might have been
concentrated upon a much smaller area -- with high per capita dose
commitments. Indeed, if the plume had carried the radio-cesiums
right to the city of Kiev, evacuation of the whole metropolitan
area might have meant visible misery for a couple of million
radiation refugees.
Instead, the fallout was spread all over Europe (European USSR
and non-USSR Europe), so per capita dose commitments are low. It
is the collective dose commitment which is huge. The resulting
radiation-induced malignancies will occur gradually and
undetectably, over many decades. They will not be distinguishable
from the very large number of spontaneous cancers occurring for
other reasons among 500,000,000 Europeans.
This one aspect of the accident-induced cancer-deaths is
emphasized very favorably by parts of the radiation community, as
we shall show.
Comments by the Authors of Nrc87:
---------------------------------
ABOUT EUROPEAN USSR -- The authors of Nrc87, Chapter 8, call
10,000 fatal cancers plus 10,000 non-fatal cancers "quite
substantial" as potential health effects from an accident (Nrc87,
p.8-10):
"The estimated effect of the Chernobyl accident on the exposed
population of 75 million is, from the standpoint of potential
health effects induced, quite substantial. Even if the Soviet
report overestimates the dose via the food pathway by an order of
magnitude, one estimates a total collective dose of about 5x10^7
person-rem. Assuming a risk factor of 2x10^-4/rem, about 10,000
fatal cancers (plus a comparable number of nonfatal cancers) would
be projected over the next 70 years."
ABOUT NON-USSR EUROPE -- The authors of Nrc87, Chapter 8,
suggest a perspective on the accident which will be frequently
echoed in other reports from the radiation community. They compare
the accident-induced dose with the unavoidable natural dose, and
the accident-induced death-rate with the entire cancer death-rate
from other causes (Nrc87, p.8-14):
"Thus, as a tentative approximation, the average individual in
Europe (outside the Soviet Union and the other countries named
above) will receive a 60-mrem dose from the accident, this dose
being spread over a period of years. For comparison, this
individual will receive about 100 mrem EACH YEAR [their emphasis]
from background radiation. Using this estimated average dose and a
total population of about 350 million people in that part of
Europe being considered, a collective dose of 2x10^7 person-rem is
calculated. Based again on a risk factor of 2x10^-4/rem, about
4000 excess cancer deaths outside the Soviet Union may be
calculated to result from the accident. These deaths would be
completely masked by the 70 million or so cancer deaths predicted
in the population over the next 70 years.
Comments by the Authors of Doe87:
---------------------------------
Return of the Threshold:
------------------------
In their Chapter 7 (at page 7.5), the authors state that there
may be a safe dose or dose-rate: "A variety of models and
assumptions can be employed in predicting possible latent health
effects in exposed populations. For example, when radiation doses
are only a few percent of natural background radiation, such doses
might be considered negligible in producing detectable adverse
health effects. For example, annual doses of 10 micro-sieverts (1
mrem), or a lifetime dose of about 500 micro-sieverts (50 mrem),
would likely produce no additional risk; thus, a major portion of
the Northern Hemisphere might produce no additional radiological
risk from the Chernobyl fallout. As noted in NCRP Report No. 64
(1980), there are no direct data that confirm that a few random
ionizations in tissue cause fatal cancers. Moreover, the BEIR
Committee noted that for low dose and dose rates, the likelihood
of zero deleterious health effects is not precluded."
Notwithstanding 1980 statements by NCRP and the BEIR-3
Committee, direct human data DO exist which confirm that random
ionizations from SINGLE TRACKS, acting independently, have caused
fatal cancers. Readers have seen the evidence themselves in
Section 5 of this book. Most of that evidence circulated widely
(as Go86) within the radiation community. The authors of Doe87 do
not refute the evidence against any safe dose or dose-rate. They
just ignore it.
Moreover, it is utterly misleading for the Doe87 authors to use
the phrase "a few random ionizations." As readers know from
Chapter 19, Part 1, the smallest possible unit of ionizing
radiation is a single primary electron track. Even for the
low-energy X-rays (30 KeV), one track from one photo-electron will
produce about (30,000 eV x one ionization per 30 eV), or about ONE
THOUSAND ionizations concentrated along its track. And not only
does single-track carcinogenesis occur -- but it might even turn
out to be overwhelmingly dominant in radiation carcinogenesis
compared with inter-track action. No one presently knows.
"Zero-Risk Model":
------------------
The Doe87 authors announce, somewhat urgently, that their
report definitely includes the threshold model in its analyses,
whereas Nrc87 -- which is NUREG-1250 -- did not. We quote (Doe87,
p.J.8):
"While NUREG-1250 does not recognize the zero risk model for
low-dose, low-LET exposure, the data do not rule out the
possibility that the cancer increase will be zero. The DOE Report,
however, contains this provision, and all cancer mortality
projections are expressed as a range, starting at zero. The zero
risk projection alternative is set forth both in the risk
projection models given in NUREG/CR-4214 (NRC 1985) and used in
the preparation of the DOE report, as well as in the BEIR report
(NAS/NRC 1980)."
And the "threshold" or zero-risk model is displayed or
mentioned everywhere throughout Doe87. We will demonstrate with
two examples.
The first is from the key table, Doe87 Table 7.11, p.7.22,
where the authors flag their column of 28,000 "Estimated Possible
Radiation-Induced Fatal Cancers with the following note: "The
possibility of zero health effects at very low doses and dose
rates cannot be excluded." The same note appears in several
tables.
The second is from Doe87 Table J.5 at p.J.7:
Location Excess Radiogenic Cancer Mortality
-------------------------------------------------------
EUROPEAN USSR 0 to 11,000
ASIAN USSR 0 to 2,500
NON-USSR EUROPE 0 to 13,000
NON-USSR ASIA 0 to 600
NORTHERN HEMISPHERE 0 to 28,000
Setting the lower end of the range at zero is a statement that
a threshold may exist, with no cancer-risk at all at doses below
that threshold.
The statement in digits and words, that Chernobyl may cause NO
cancer-deaths, is made so many times in Doe87 that we lost count.
It is mentioned four times even in the "Executive Summary."
Comparison with Entire Cancer Problem:
--------------------------------------
Also more times than we can count, Doe87 makes the comparison
between the 28,000 "estimated possible radiation-induced fatal
cancers" and the entire number of cancers which will occur anyway.
It starts in the Executive Summary.
The table on page xii tabulates spontaneous and
radiation-induced fatal cancers side by side. Among 3.5 billion
people in the Northern Hemisphere, Doe87 expects about 600,000,000
"natural" fatal cancers (about one death in six), and lists 28,000
Chernobyl-induced fatal cancers -- annotated with the speculation
that the possibility of zero cancers "cannot be excluded."
On the next page, the text makes the comparison in words:
"Estimates of excess cancer cases, which may be as low as zero for
the majority of exposed populations, are so small that they are
negligible compared to the higher cancer mortality from natural or
spontaneous causes in those populations" (Doe87, p.xiii). The
28,000 possible Chernobyl-induced deaths are described as a
possible 0.004 percent increase in cancer-mortality (p.xiii,
p.7.22). The percent is (28,000 / 600,000,000) x (100), of course.
Comparison with Natural Dose:
-----------------------------
Another recurring theme, in the authors' own comments, is the
comparison of per capita dose commitments from Chernobyl with the
dose commitment received by humans from natural background
sources. One example suffices. Discussing 50-year dose-commitments
in Non-USSR Europe, Doe87 says (p.5.62):
"... the calculated average dose commitment to the population
of any listed country is less than 5 mGy (500 mrad). Thus,
although the calculated total collective dose commitment is large,
the average individual dose commitments for even the European
countries are equivalent to that received from background
radiation in a few years."
Comments by the Authors of Doe88:
---------------------------------
At the Beginning:
-----------------
The abstract of the SCIENCE version is very brief, and features
this statement: "The best estimates for the lifetime expectation
of fatal radiogenic cancer would increase the risk from 0 to 0.02
% in Europe and 0 to 0.003 % in the Northern Hemisphere" (Doe88,
p.1513).
Immediately following the abstract are three introductory
paragraphs in which Anspaugh, Catlin and Goldman describe the
Chernobyl accident as "the largest reported accidental release of
radioactive material." They wish to put this into perspective:
"The purpose of this article is to present a global perspective
of the significance of the release." They add, "The dominant
concern for the world's citizenry after the Chernobyl accident has
been future risks to health. This concern continued even after it
was clear that the individual risks outside the Soviet Union would
be quite small," at which point they cite their own DOE 1987
report.
"Chickens Come Home to Roost":
------------------------------
Why did the public continue to be concerned in spite of the
reassuring report from DOE in 1987?
In the Preface of the 1987 report, Goldman, Catlin, and
Anspaugh describe themselves and the co-authors as dedicated
scientists: "A dedicated group of scientists from 11 research
institutions [mostly DOE-funded laboratories] have contributed to
making this report possible ... Many of the models and values
chosen for parameters used in this report stem from research that
has been sponsored by the U.S. Department of Energy. The spectrum
of such radiological health and environmental research over the
past 4 decades includes... pioneering advances in risk assessment"
(Doe87, p.vi). And Doe87 was the mother of Doe88.
These authors seem unaware that DOE reports have no credibility
at all with much of the public, in view of DOE's inherent conflict
of interest coupled with its record of covering-up the careless
radioactive contamination around many of its own facilities and
its record of other problems.
Indeed, soon after Doe88 -- and following pressure from citizen
lawsuits, FBI investigations, and the prospect of criminal
prosecution of some DOE employees -- Energy Secretary James
Watkins would admit in June 1989:
"... the chickens have finally come home to roost, and years of
inattention to changing standards and demands regarding the
environment, safety and health are vividly exposed to public
examination, almost daily. I am certainly not proud or pleased
with what I have seen over my first few months in office"
(Wat89a).
And even more recently, Watkins is still expressing dismay over
DOE performance. Referring in December 1989 to DOE's plans for
waste burial in Nevada and New Mexico, he said that "the whole set
of schedules was not scientifically sound, not fiscally sound, not
technically sound... They were incomplete, misleading, and not
properly done" (Wat89b).
On the problem of candor, Watkins said that DOE will soon issue
rules to protect lower-level employees who make allegations about
safety, competence or the honesty of their superiors. "We've been
totally unresponsive to whistle-blowers," Watkins said (Wat89b).
In the Middle:
--------------
In Doe88, between its beginning and its end, the authors assert
ten times in six pages that there may be zero Chernobyl-induced
cancers. As justification, they say only, "We have taken the
bottom of the range [of cancers] to be zero, which is consistent
with the NUREG report" (Doe88, p.1515.)
There is a lack of symmetry here. If Anspaugh, Catlin and
Goldman wish to stress NUREG's absolutely lowest risk at every
opportunity, they are scientifically obliged to give equal
emphasis to NUREG's so-called "upper bound estimate" (from the
linear model). They quantify it only once in their summary
(p.1518), as quoted below, and they do not show that NUREG's
"upper-bound" risk-factor would increase the Doe88 estimate of
17,400 Chernobyl-induced cancers by about 3-fold, to at least
50,000 fatal cancers.
In Their Summary:
-----------------
Anspaugh, Catlin, and Goldman provide a summary of the "global
perspective" as follows:
"Outside of the immediate Chernobyl region, the magnitude of
radiation doses to individuals is quite small, leading to
extremely low incremental probabilities of any person developing a
fatal radiogenic cancer over a lifetime ... Probably no adverse
health effects will be manifest by epidemiological analysis in the
remainder of the Soviet population [outside the immediate
Chernobyl region] or the rest of the world. Projections of excess
cancer risk for the Northern Hemisphere range from an incremental
increase of 0 % to 0.003 %. An upper bound estimate would range
from 0 % to about 0.01 %, still undetectable ... The social
consequences are more difficult to quantify, but public concerns,
whether justified or not, have increased, necessitating attention
by medical, public health, and other authorities" (Doe88, p.1518).
Their perspective has the familiar format -- many more people
will NOT be killed than WILL be killed. Perhaps a global
perspective is adaptable for Bhopal, famine, World War Two, or
even homicide.
If a "global perspective" is considered today, why not an
inter-stellar perspective tomorrow? With a bit more advance in the
space program, we will find out how many other places support
life, and then someone can estimate the INTER-STELLAR impact of
nuclear accidents which occur on Earth ... and the inter-stellar
impact will surely be much smaller than the global impact.
9. The Threshold and Dose-Exclusion: Ultra-Low Cancer Estimates
================================================================
It is undeniable that the Chernobyl accident has made the
concept of a safe dose or dose-rate more attractive than ever. It
is understood that 17,400 to 475,000 cancer-deaths from a single
accident do not provide a fertile ground for the nuclear
enterprise, which funds (via its governmental and private arms)
most radiation research worldwide. A perception of ZERO
cancer-deaths would be much more favorable.
Can this need for a threshold be met on a scientific basis?
Having presented our DISPROOF of any threshold, we answer "No," of
course. But elswehere, as we have already shown, one may face
temptation to presume a threshold, without having any appropriate
basis in science and without even dealing with the conclusive
evidence AGAINST it. Under such a presumption, the Chernobyl
problem could be "solved" by throwing away about 95 percent of the
collective dose commitment, because it would lie BELOW the
presumed threshold. Handling scientific issues in such a manner
would be truly a "crossroad in the radiation health sciences."
Some ultra-low Chernobyl estimates follow.
An Article in the Official Journal
of the Society of Nuclear Medicine
-----------------------------------
The first article we will examine is by a member of all the key
BEIR-3 Committees (see Chapter 37): Edward W. Webster, Ph.D.,
Department of Radiological Sciences, Massachusetts General
Hospital.
The article is entitled "Chernobyl Predictions and the Chinese
Contribution," in the April 1987 issue of THE JOURNAL OF NUCLEAR
MEDICINE (Webs87). It is based on a paper given on November 6,
1986.
Webster begins by calling Chernobyl-induced cancers "obviously
speculative" and offering a perspective of his own: "As of this
writing, the only certain effect has been the 31 early deaths, and
therefore to-date the casualties are much smaller than the
hundreds who died in each of the several recent crashes of jumbo
jet aircraft, and the thousands who died in the chemical disaster
at Bhopal, India" (Webs87, p.423).
He goes on to point out, correctly, that the issue of predicted
cancers will interest nuclear medicine physicians since individual
doses from the accident are typically "well below those
administered in diagnostic nuclear medicine" (Webs87, p.423).
Also correctly, Webster states: "The predictions cover a wide
range, heavily dependent on the assumptions made concerning the
relation of cancer to low-level radiation exposure, and somewhat
less dependent on dose assessments. At the high end of the range
are those of John Gofman, PhD, MD," and he cites my estimate given
at the American Chemical Society meeting (Go86).
Webster continues: "Dr. Gofman's prediction is unique insofar
as it employs his own estimate of lifetime cancer risk per rem,
whereas most other predictions utilize the risk estimates adopted
by the International Commission on Radiological Protection (ICRP),
the United Nations Scientific Committee on the Effects of Atomic
Radiation (UNSCEAR), and other international bodies."
By definition, independence from the official line is the
ESSENCE of an independent analysis -- although independence alone
does not make the analysis CORRECT, as we pointed out in Chapter
2.
In the same chapter, we also pointed out a set of circumstances
(chiefly funding) which can produce an ARTIFICIAL consensus of
experts. It is interesting to contrast the views of Dixy Lee Ray,
a former head of the U.S. Atomic Energy Commission (AEC), with the
predicament of James Watkins, current head of AEC's replacement,
DOE. Moghissi and Ray (Mog89) have been insisting that consensus
in science means everything, whereas Watkins is finding that
experts can be persuaded to reach a consensus on managing
radioactive waste which is "not scientifically sound ...
misleading, and not properly done" (see Part 8).
Webster's First Recommendation:
-------------------------------
Webster continues, still correctly (Webs87), p.423: "The Gofman
risk estimate... is about 40 times higher than the above
`official' estimate of 100 [cancer-deaths] per million
person-rems" -- which is a Cancer-Yield of 1.0. Webster appears to
prefer the 1977 UNSCEAR value to the higher BEIR-3 value of about
2.0.
Webster makes two recommendations for resolving the disparity
between my estimates and `official' estimates, and for estimating
Chernobyl-induced cancers.
First, he suggests that the world will find out the "correct"
value for Cancer-Yield from an epidemiologic follow-up of 24,000
highly exposed persons near the Chernobyl explosion (Webs87,
p.424).
By contrast, we strongly caution against any policy which would
make the radiation health sciences and human health itself
dependent, in any measure at all, upon Soviet data on a radiation
issue. Readers are referred back to Part 1 of this chapter.
Webster's Second Recommendation:
--------------------------------
Webster also recommends that meanwhile, the radiation community
should give great weight to a recent "Denver-Type" study (our
term, not his) from the People's Republic of China, in order to
resolve the disparity in Cancer-Yields: "The Gofman estimate
appears particularly improbable in the light of the Chinese study"
(Webs87, p.425).
The study to which he refers is by Zufan and Luxin (Zu86)
entitled "An Epidemiological Investigation of Mutational Diseases
in the High Background Radiation Area of Yangjiang, China," in the
JOURNAL OF RADIATION RESEARCH (JAPAN). Zufan and Luxin thank RERF
in Hiroshima for editorial assistance. Luxin is a Chinese delegate
on UNSCEAR-88.
The study (Zu86) finds the cancer-rate in the high background
area to be lower than the cancer-rate in the low background area.
The paper is one of several earlier and later Chinese reports on
their high background area (see also Chapter 35, Part 7).
This type of study is inherently unable to resolve anything
about the low-dose and threshold issues, as explained in Chapter
21, Part 2. The BEIR-3 Committee made much the same point with
respect to some earlier "Denver-Type" studies (Beir80,
pp.469-471). Therefore we are critical of reliance on this study
and of its representation to the physicians as a key study.
The Need for Proper "Blinding":
-------------------------------
Moreover, another aspect of the paper by Zufan and Luxin
deserves attention. The study may have an open doorway for bias to
confound its results. The authors state the following (Zu86,
p.143):
"Cancer mortality in the high background radiation area and the
control area has been investigated for more than 14 years. The
early data (1970-1978) were obtained by means of a retrospective
survey. In 1979, a cancer registry system was established for the
study areas whereby local physicians, with the help of many
hospitals and administrative organizations, report all incident
cancer cases and cancer deaths to the registry. Diagnoses are
confirmed by an expert group who meet to evaluate cases once or
twice a year."
In other words, this is not even a Denver-Type study based on
Vital Statistics compiled by persons with no knowledge of the
study. In the Zufan study, the statistics are first collected with
a well-known purpose, and then re-evaluated by an "expert group"
with full knowledge of the purpose.
The opportunities for bias to enter are self-evident. The paper
mentions not even one precaution against such bias. If input to
the study's database were to include some over-diagnosis of cancer
in the low background area, or some under-diagnosis in the high
background area, the study's output -- its "answer" -- could be
easily pre-determined at the outset.
We are disappointed that peer-reviewers did not insist that the
"blinding" problem be shown as solved, or be acknowledged if NOT
solved.
Embracing Data from China:
--------------------------
Ideally, a scientific report deserves to stand or fall on its
own merits, and not because of its source. We made that point
emphatically at the end of Chapter 2.
But also one is obliged to be realistic about the misuse of
science in the service of policy. (See also warnings by Dr.
Sheldon Wolff on this same subject, in Chapter 35, Part 4).
As stated with regret in part 1 of this chapter, I warn against
acceptance of uncheckable data or findings coming out of any
country whose authorities have recently or currently demonstrated
no regard for truth when it undermines policy.
In China, the policy has been to undertake nuclear power
generation. By 1982, plans were underway to build such plants just
north of Hong Kong (Nyt82). And the policy has been pursued
against popular protest -- a million signatures in Hong Kong
against it, according to the WALL STREET JOURNAL of April 13, 1987
(Wsj87). Under the circumstances, it is common sense to say that
the government would welcome reports suggesting that a little
radiation is harmless or possibly even good for people.
It is realistic to worry that radiation analysts in the
People's Republic of China -- especially in the absence of a free
press there -- may expect to pay a heavier price than radiation
analysts elsewhere, if they were ever to question whether data
sponsored by the state (on background doses, cancer
mortality-rates, or anything else) were rearranged, falsified,
selectively abbreviated, or just plain fabricated. Individual
analysts, fully innocent themselves, could be deceived under such
regimes without even knowing it for certain.
100 Chernobyl-Induced Cancers:
------------------------------
Notwithstanding all these problems, Webster looks very
favorably on what his title calls "the Chinese Contribution." His
article ends as follows:
"The Chinese evidence at present suggests that the excess
cancer mortality from the long-term exposure to low levels of
external and internal radioactivity of many millions in Russia and
Europe could be less than 100 and is almost certainly below a few
thousand. The Chinese contribution to our knowledge of low-level
radiation is still developing, and the present provocative
findings may change or may reveal an explanation which will admit
support for current risk estimates. Potentially, the impact of a
larger statistical study with a zero or negative index of
low-level radiation effect could be very far reaching."
Nowhere do the nuclear physicians receive warning about the
inherent limits of Denver-Type studies, about the "blinding" issue
in this study, and about the even bigger issue of caution toward
unverifiable reports from certain nations.
These physicians may infer, mistakenly, that the "Chinese
contribution" is valid evidence in favor of a safe dose -- an
inference which could have unintended consequences for their
patients and staffs. Webster himself must assume a safe dose when
he suggests Chernobyl-induced cancers "below a few thousand" or
even "less than 100." As for the conclusive evidence AGAINST any
safe dose -- presented in Go86, which Webster cites -- Webster
does not refute it or even mention it.
An Article in the Official Journal of the Health Physics Society
----------------------------------------------------------------
The second article we will examine is by Robert E. Alexander.
He is the 1988-89 President of the Health Physics Society.
Elsewhere, he identifies himself also as (A) a scientist with the
U.S. Nuclear Regulatory Commission, and (B) a member of the
Science Panel preparing a report for the Veterans Administration
to "assist in the adjudication of claims of service-related
radiogenic cancer" (Alex88a, p.145; Alex88b, p.592).
The article is entitled "A New Intellectual Atmosphere," in the
June 1988 issue of HEALTH PHYSICS, which describes itself as "the
radiation protection journal" on its cover. This article (Alex88b)
is a guest editorial. Sections of this article also appear in a
much shorter article entitled "Health Effects from Radiation," in
the February 1988 issue of ENVIRONMENTAL SCIENCE & TECHNOLOGY
(Alex88a).
Concern about "Decision Makers":
--------------------------------
Alexander is quite forthright about the importance to the
nuclear enterprise of shaping the perception of Chernobyl's cancer
consequences -- especially the perception of decision-makers:
"... predictions of delayed deaths from radiation-induced
cancer seem to me to be the most significant reactor accident
consequences in terms of impressions left with decision makers. I
suspect it is these estimates that are more likely to prompt the
word `catastrophic' and to alarm decision makers" (Alex88b,
p.589).
"... very small doses to very large numbers of people can yield
very alarming results" (Alex88b, p.592), at which point he cites
the Doe87 estimate of 28,000 Chernobyl-induced cancer-deaths.
Then he calls the Doe87 estimates for European USSR and for
Non-USSR Europe "conjecture, i.e., inference from insufficient
evidence and not useful for decision making" (p.592). On the next
page and also in the shorter article he says:
"In my opinion there is a very limited place for conjecture and
speculation in science. Even hypotheses must always be clearly
identified as such, particularly when the results of hypothetical
calculations can reach unsuspecting legislators and agency heads,
influencing their decision-making process in a manner detrimental
to the best interests of the nation" (Alex88a, p.145; with minor
differences, Alex88b, p.593).
"There is a larger picture that should be considered. The
catastrophe that I am worried about is that the energy needs of
many people may be delayed by those who fear that the sky is
falling" (Alex88b, p.593).
Speculations about a Threshold
------------------------------
Alexander recognizes, as everyone must, that acceptance of
nuclear energy would be vastly easier if there were acceptance of
a threshold.
In support of the threshold hypothesis, he cites (Alex88b,
p.592) a number of Denver-Type studies and the A-Bomb Study
1950-1978. We have already explained why all of these studies are
inherently incapable of answering the threshold question, however.
Alexander does not refute or even mention the conclusive and
appropriate human evidence AGAINST any safe dose. Mostly, he
relies upon quoting threshold allusions from Doe87 and from the
1980 BEIR-3 Report (provided, respectively, to readers by us in
Part 8 of this chapter and in Chapter 34).
The threshold speculation is competing with good human
evidence. When the speculation about upward curvature for human
dose-response was competing with good human evidence, the
speculation prevailed. If the threshold speculation prevails, then
95 percent of the dose commitment and consequences from Chernobyl
can be thrown out.
Goldman, Catlin, and Anspaugh appear to have been the pioneers
in this -- which is consistent with their description of DOE as
the sponsor of "pioneering advances in risk assessment" (Doe87,
p.vi). Although Doe87 made its range of Chernobyl-induced cancers
"0 to 28,000," it also explored dose-levels at which a threshold
would be important (Doe87, p.5.46):
"Another question of interest is how much the total collective
dose might be reduced if the calculation were made with the
exclusion of very small, but nonzero, individual total-body doses
of, for example, less than 0.5 mGy (50 mrad)." The authors report
that with this exclusion, "... the calculated total collective
dose commitment would decrease by less than 6% ."
So it would seem that a speculative threshold at 50 millirads
cannot solve the Chernobyl problem. In the next paragraph, the
authors explore 500 millirads:
"To put these dose estimates into further perspective, it
should be noted that if individual lifetime dose commitments below
5 mGy (500 mrad) are excluded, all but the more heavily affected
portion of the USSR would be removed from the global collective
dose summary" (Doe87, p.5.46).
Now this could be USEFUL threshold information.
410 Chernobyl-Induced Cancers:
------------------------------
And it is soon used. On page 7.8, Goldman, Catlin, and Anspaugh
suggest that the population evacuated from the 30 kilometers
around the former reactor will experience between zero and 410
cancer fatalities from their external exposure. (Doe87 used the
Soviet estimate of 135,000 evacuated persons in this context; the
Soviets reduced the number to 115,000 persons before Doe88.)
The number "410" is picked up by Alexander and featured in both
his long and short articles (Alex88b, p.591 quoted below;
abbreviated in Alex88a, p.145):
"Consider the example of the 28,000 cancer death estimate for
Chernobyl. If individual doses below 0.1 Gy (10 rads), and dose
rates below 0.01 Gy y^-1 (1 rad y^-1) lifetime, are excluded from
the calculation, only the evacuees are affected and the
theoretical result is 410 cancer deaths. A difference of this
magnitude is sufficient to alter conclusions." Indeed.
10. Beyond Chernobyl: The Much Bigger Agenda
==============================================
Chernobyl is only "the tip of the iceberg" with respect to the
concept of dose-exclusion. There is a bigger agenda under
discussion, and Alexander's article serves as one illustration.
Alexander makes it clear, by his own words below, that he
disapproves in GENERAL of including individual doses below 10 rads
and dose-rates below one rad per year in current risk-benefit
considerations. Those levels are the ones below which the BEIR-3
Report declined to quantify risk coefficients (our Chapter 34),
even though its own analysis of solid cancer in the A-Bomb Study
produced a linear dose-response. Alexander writes:
"It is understandable that many health physicists are dismayed
by the now common practice of including extremely low doses in
collective dose calculations. When doses obtained in this manner
are multiplied by risk coefficients, valid at best for doses and
dose rates exceeding those specified by the BEIR-III Committee,
the results can be alarming, misleading and they may have
detrimental influence on decision makers" (Alex88b, p.591).
After telling readers that the Nuclear Regulatory Commission is
proposing to establish a "de minimis" dose of one millirad for
collective dose calculations, Alexander says that the
Environmental Protection Agency is opposing the NRC proposal. He
blames the behavior of EPA and "government officials" on their
ignorance:
"It is inconceivable to me, to mention only three examples,
that government officials actually aware of the assumptions made
in connection with low-level radiation risk assessments would have
(1) approved $2 billion for decommissioning of formerly used U.S.
Atomic Energy Commission (AEC) facilities and other aspects of the
DOE Remedial Action Program, (2) established NRC effluent-control
design criteria of ...8 mrem y^-1 for nuclear power plants or (3)
taken the U.S. Environmental Protection Agency (EPA) position that
Environmental Impact Statements using non-zero lower limits of
collective dose integration are not acceptable" (Alex88b, p.593).
"Reasonable people will not knowingly want to support proposals
for large expenditures to protect against risks that have an
entirely theoretical basis, that may not exist, and that can never
be demonstrated (Alex88b, p.594).
"The nation is expending enormous resources to protect the
public against risks believed by an overwhelming, but silent,
majority of the scientific community to be trivial or even
non-existent" (Alex88b, p.594).
"Below Regulatory Concern":
---------------------------
Silent? Regulatory bodies seldom respond to silence, and yet
proposals are moving forward in the U.S. Nuclear Regulatory
Commission to declare a large share of radioactive waste to be
"below regulatory concern" and to treat it just like
non-radioactive waste in local landfills, incinerators, sewage
plants, and recycling circles.
So Sorry If We're Wrong . . .
-----------------------------
Some segments of the radiation community appear to believe
passionately that no one should impede the nuclear enterprise on
the basis of what they label as speculation and conjecture about
injury from low doses and dose-rates. Instead, they ask the world
to accept THEIR speculation and conjecture that low doses and
dose-rates are safe -- a notion which would surely result in
increased exposures.
But if the threshold speculation is wrong (as shown in this
book), and nonetheless we contaminate the planet irreversibly with
radioactive poisons, the results might be hundreds of millions of
unnecessary cancers over time -- as well as a presently
unquantifiable price in heritable genetic damage.
Price of Past Presumption:
--------------------------
Society has acted before, in previous decades, on the basis of
rosy but mistaken presumptions promoted by parts of the radiation
community.
In the absence of conclusive evidence, optimistic assumptions
in this field have led to past "benefit-risk judgments in which
the benefit was sometimes real, but the cancer-risk from the
associated doses was casually dismissed. Today some of the
practices in the list below continue, but usually at much lower
doses than in the past. The following is merely a partial listing:
* Use of luminous radium dials in wrist-watches and airplane
instruments (chronic gamma irradiation of the cockpit crew).
* The promotion of radon spas and radium-laced water as
health-enhancers.
* Use of fluoroscopy machines in shoe stores, with some
unavoidable dose not only to the pelvis, but also to the face
and neck of people looking down to enjoy the sight of their
foot-bones.
* Use of the fluoroscope by voice teachers to show the position
of the diaphragm at the beginning, middle, and end of a
singer's phrase.
* Irradiation of infants in utero during maternal pelvimetry.
* Routine irradiation of infants for a "disease" (thymic
enlargement) which was later admitted never to have needed
any treatment at all.
* Routine irradiation of tuberculosis patients to monitor
pneumothorax treatment.
* Irradiation of women for post-partum mastitis.
* Irradiation of people for ringworm of the scalp.
* Cobalt treatment for blocked eustachian tubes.
* Radium treatments for "sinus trouble."
* Use of X-ray exams to monitor the advance or regression of
curvature of the spine (scoliosis), mostly in young girls.
* Fluoroscopic exams of babies as part of routine "check-ups."
* Use of radioactive thorotrast as a routine contrast medium in
diagnostic radiography.
* The practice of giving full-spine X-rays, "GI series" and
barium enemas as part of the routine "annual check-up" in the
1940s.
* The smoking of cigarettes whose tobacco-smoke is contaminated
by radioactive decay-products from uranium, present in the
soil or in phosphate and raffinate fertilizers.
* The use of young nurses and young mothers to hold small
children during X-ray exam of the child.
* Absence of lead-shielding between X-ray offices and adjacent
offices and elevators.
Several of these past practices provided the early
epidemiological proof that ionizing radiation can induce fatal
human cancers.
One needs to wonder seriously how much of the current
cancer-rate is due to past exposure to ionizing radiation from
such practices. It could be a meaningful part of the so-called
"spontaneous rate.
"De Minimis" -- Beyond Chernobyl:
---------------------------------
"De minimis non curat lex", or "the law does not concern itself
with trifles," is referred to simply as "de minimis" in proposals
NOT to count a certain amount of population exposure from ionizing
radiation in risk-analysis -- and NOT to regulate certain amounts
of radioactive pollution. Of course, the two issues are closely
related to each other.
The most extreme position, probably supported by very few in
the radiation community, favors the exclusion from
risk-considerations of all individual doses when the individual's
risk is small, regardless of the magnitude of the COLLECTIVE dose.
This is another way of saying that even 950,000
Chernobyl-induced cancers would not be worth attention, because --
although the collective dose and health-price might be huge --
each INDIVIDUAL's dose and risk would be very small. If this type
of "de minimis" proposal ever prevails, the health consequences
from Chernobyl-size accidents (or the equivalent from gradual
PLANNED emissions) could be officially treated as negligible.
Less drastic "de minimis" proposals would give some
consideration to the magnitude of the collective dose -- with some
limit on the person-rads per source which would not "count." Of
course, if sources were subdivided into regions, facilities, or
ultimately into particular vents or pipes, the true collective
dose "not counted" could become larger and larger.
"De minimis" proposals are a "hot" topic, and certainly NOT
everyone in the radiation community supports the concept. Whatever
decisions are made, it seems safe to predict that policies
accepted in the radiation health sciences will influence policies
set in other health sciences, too.
We will quote Bo Lindell of Sweden's National Institute of
Radiation Protection. He is also a member emeritus of the ICRP's
Main Commission, and is a Swedish delegate to UNSCEAR (see Chapter
37). In a thoughtful letter to HEALTH PHYSICS, he concludes
(Linde89):
"... I suggest that the profession of radiation protection
should adopt a cautious attitude rather than belligerently crying
for a de minimis, a concept which I consider untenable on both
logical and ethical grounds."
"De Minimis" -- Beyond Radiation:
---------------------------------
Many people have observed that human nature incorporates some
contradictory tendencies. It seems contradictory to me that, on
the one hand, there is a readiness to inflict cancer-death on
undetectable victims who will not be noticed, while there is a
competing tendency which causes some people in Oakland,
California, to risk their own lives on an unstable structure and
work themselves to exhaustion following the October 1989
earthquake, just on the very slim chance that they might SAVE one
life from under the collapsed freeway.
People of goodwill need to look closely at the aggregate
consequences of individually small risks. If pollution sources of
all types are regulated individually, and each is allowed under
the "de minimis" concept to kill one person in 100,000 (a low
individual risk), then only 10,000 sources could kill up to one
tenth of the population. And no one would ever be able to prove
it.
A Reality-Check on Confidence:
------------------------------
When various experts advocate that we neglect to "count" or
evaluate exposure to some pollutant below an arbitrary dose or
dose-rate, they generally claim that the low dose or dose-rate
will be too trivial to matter: "A smaller hazard than getting out
of bed." Thus such experts should not object to pre-testing their
own proposals before scaling them up to everyone.
After all, if the proposed doses are such a trivial hazard that
the experts say the general public should not object, then why
should these same experts object to exposing their OWN children
and grandchildren intentionally to all the proposed doses, for the
next 10 to 20 years?
I wonder if such guardians of the public's health might think
twice, before agreeing to a personal kind of pre-testing for their
policies -- BEFORE they are applied to children everywhere.
--------------------------------------------------
============================================================================
----------------------------------------------------------------------------
Table 24-A
Comparison of Chernobyl-Induced Fatal Cancers, Estimated by Gofman, NRC, and
DOE.
======================================================================================
| Col.A | Col.B | Col.C | Col.D |
| | | | |
| | Whole-Body | Fatal Cancer-Yield | Chernobyl-Induced |
| | Dose Commitment | (Fatal Cancers per | Fatal Cancers |
|Source of Estimate: | in Person-Rad | 10,000 Person-Rad) | (estimated) |
|====================================================================================|
| | | | |
|Gofman Sept. 1986. | 127.4 million | 37.313 | 475,500 |
| Part 3 of | person-rad. | | |
| this chapter. | All-time commitment. | | |
|------------------------------------------------------------------------------------|
| | | | |
|NRC January 1987. | 70 million | 2.0 | 14,000 |
| Part 4 of | person-rad. | | |
| this chapter. | Fifty-year cut-off. | | |
|------------------------------------------------------------------------------------|
| | | | |
|DOE June 1987. | 120 million | 2.3 | 28,000 |
| Part 5 of | person-rad. | | |
| this chapter. | Fifty-year cut-off. | | |
| | --------------------------------------------------------------|
| | | | |
| | If Doe87 had used | | |
| | the corresponding | | |
| | all-time commitment: | | |
| | | | |
| | 150 million | 2.3 | 34,500 |
| | person-rad. | | |
|------------------------------------------------------------------------------------|
| | | | |
|DOE December 1988. | 93 million | 1.87 | 17,400 |
| Part 6 of | person-rad. | | |
| this chapter. | Fifty-year cut-off. | | |
| | --------------------------------------------------------------|
| | | | |
| | If Doe88 had used | | |
| | the corresponding | | |
| | all-time commitment: | | |
| | | | |
| | 116 million | 1.87 | 21,700 |
| | person-rad. | | |
| | | | |
======================================================================================
There is no mystery about the disparities in Column D.
These differences cannot be blamed on the relatively small differences in
estimated dose. Indeed, the Gofman and DOE estimates are remarkably close.
Part 6 explains why DOE needs to use the all-time dose commitment -- not the
50-year cut-off.
The differences in the estimated Chernobyl-induced cancers lie
overwhelmingly in an independent evaluation of Cancer-Yield (cancer-risk)
versus the Cancer-Yields used by the radiation community.
----------------------------------------------------------------------------
============================================================================
============================================================================
----------------------------------------------------------------------------
Table 24-B
A Realistic Range for Chernobyl-Induced Fatal Cancers, Based on Gofman and
RERF.
----------------------------------------------------------------------------
Every estimate here is based on a collective all-time dose commitment of
127.4 million person-rad. This value (from Go86) lies between the DOE
all-time dose commitments of 116 and 150 million person-rad. See Table 24-A,
Column B.
========================================================================================
| Col. A Col.B Col.C Chernobyl-Induced |
|Source of the Estimate of Lifetime Fatal Lifetime Fatal Cancer Fatalities |
|Fatal Cancer-Yields Cancer-Yield Cancer-Yield |
| T65DR Dosimetry DS86 Dosimetry T65DR DS86 |
|======================================================================================|
| | |
|Gofman Cancer Difference Method. | |
|A-Bomb Study, 1950-1982. | |
|Low-Dose Exposed vs Ref. Grp. | |
|Table 14-C, Row 1. 16.2 12.23 | 206388 155810 |
|-----------------------------------------------------------------+--------------------|
|Gofman Cancer Difference Method. | |
|A-Bomb Study, 1950-1982. | |
|Best Fit by Regression. | |
|Table 14-C, Row 2. 12.9 12.03 | 164346 153262 |
|-----------------------------------------------------------------+--------------------|
|Gofman Cancer-Rate Ratio Method. | |
|A-Bomb K-values and A-Bomb Survivors. | |
|Table 16-B. 31.65 30.43 | 403221 387678 |
|-----------------------------------------------------------------+--------------------|
|Gofman Cancer-Rate Ratio Method. | |
|A-Bomb K-values and U.S. Population. | |
|Table 16-C. 26.64 25.56 | 339394 325634 |
|-----------------------------------------------------------------+--------------------|
|RERF: Shimizu and co-workers. | |
|Sub-cohort, A-Bomb Study, 1956-85. | |
|Table 19, page 53, Shi88. | |
|Details in our text, Part 7. NOT DONE 12.4 | NOT DONE 157976 |
|-----------------------------------------------------------------+--------------------|
|RERF: Preston and Pierce. | |
|Sub-cohort, A-Bomb Study, 1950-85. | |
|Pr88, page 458. | |
|Details in our text, Part 7. NOT DONE 11 | NOT DONE 140140 |
|-----------------------------------------------------------------+--------------------|
|Gofman: Worldwide Low-LET Human | |
|Evidence, with Variable Rel. Risk. | |
|A-Bomb Study, 1950-1974 included. | |
|Go81. 37.313 NOT DONE | 475368 NOT DONE |
========================================================================================
The entries in Column D for the T65DR dosimetry are (127.4 million x
Col.B) / (10,000).
The entries in Column D for the DS86 dosimetry are (127.4 million x
Col.C) / (10,000).
----------------------------------------------------------------------------
The Cancer-Yields in Columns B and C are central estimates based on the
best available real-world human evidence. By contrast, Cancer-Yields in the
range of 1.0 to 2.3 used by the radiation community are grossly at variance
with this evidence, as we demonstrated in this book by showing step-by-step
what really does emerge from the evidence.
If DOE, for instance, would just use reality-based Cancer-Yields instead
of Cancer-Yields based on preferred speculations, the disparity among
estimates of Chernobyl-induced cancer-deaths would shrink to about
three-fold, as shown in Column D above.
----------------------------------------------------------------------------
Each entry in Column D needs doubling, if one wishes to include non-fatal
cancers. Table 24-B shows that scientifically reasonable estimates of
Chernobyl-induced cancers range in the hundreds of thousands, not the tens
of thousands.
----------------------------------------------------------------------------
============================================================================