The terms representative persons (for example, in the abstract and on page 2) and representative individuals (as on page 57) should be avoided; they have connotations that some person—in fact, maybe most people—actually received the dose in question.
The term population weighted needs clarification and definition.
It should be made clear that the uncertainty of the fission-to-fusion ratio most likely led to an overestimate of the dose.
Change “designated” to “detonated”.
Give a reference.
Page 39, last paragraph
“Therefore, the doses that are presented in this report are those arising from the decay of gammaemitting radionuclides deposited on the ground…” Such references to dose give the false impression that doses are well known (that is, are without significant error). It is recommended that references to dose in the report be indicated as “dose estimates” where appropriate.
There is confusion related to the subscript R. As the equation is written, there is a need for a different subscript for E, perhaps Ei to refer to the “effective dose from radionuclide i”. Then the subscript R can be used with w as specified by ICRP as the radiation-weighting factor. This change is required because a radionuclide can emit radiation with different weighting factors. With those suggested changes, the equation will correctly read that the effective dose of a given radionuclide is summed over all radiation and tissue-weighting factors.
“If further work is conducted, it should include 239Np in the internal dose calculations.” What is the basis for this recommendation? What fraction of internal dose has been omitted because 239Np is omitted here?
Page 46, second paragraph
It is stated that “the total deposition of Cs-137 from all Nevada Test Site tests considered in this report through 1962 is shown in Figure 3.1.” The information presented in the figure is displayed as deterministic (that is, without any error or uncertainty). It is recommended that the uncertain nature of the information presented be mentioned.
“For most of the country, the amount of cesium was 10 to 20 times the amount of plutonium deposited.” That must mean on an activity basis. It should read, presumably, “…the activity of cesium was….” In addition, Table 3.5 indicates that the activity ratio applies for 239+240Pu, not including 241Pu.
“…all Nevada Test Site tests, 34% of the 137Cs produced was deposited in the contiguous United States.” Were similar calculations made for the other nuclides? It might be useful to make a similar observation for the other radionuclides, and comment (if necessary) on the differences between the fractions deposited for each radionuclide.
Page 50, Figure 3.5
Results for the fraction of total 137Cs deposited in the United States from the NTS by year of test are presented as exact values (without error). It is recommended that the results be more appropriately described as approximations. This recommendation also applies to other inappropriate deterministic statements in Chapter 3 related to calculated results.
Page 51ff, Section 3.2.2
The discussion in this section relating exposure, external dose, effective dose, and absorbed dose is confusing. It switches illogically between exposure and dose, with non sequiturs. There is also a disconnect between the radionuclide-weighting factors of Equation 3.1 (not discussed in the text) and the radiation-weighting factors discussed later in the text.
“The conversion factors relating deposition density to exposure rate in air have been validated in many studies and are believed to be accurate to within 5% (NCRP 1999).” That statement must apply to situations where the deposition is on well-characterized surfaces. For deposition on surfaces in general, the uncertainty is probably higher.
It is recommended that where effective dose is introduced it be pointed out that the effective dose is a hypothetical dose in that it is based on use of the hypothetical linear no-threshold risk model.
Page 54, lines 18–20
The authors should note that this relationship applies only to low-LET radiations.
Page 54, first full paragraph
This paragraph is confusing, coming just after a discussion of the relation of effective dose to absorbed dose (Equation 3.1 and following text). The reader is given an approximate relation between exposure and effective dose, and then external dose related to effective dose (“numerically equal”).
Page 54, first full paragraph
Does this mean the roentgen-to-sievert conversion is 0.0066, or should it be 0.0093?
The section on external dose should be rewritten for clarity. There is an amalgamation of SI and traditional units, which is confusing.
The discussion indicates that an assumption of about 80% indoors leads to an effective dose of about 44% of the outside dose. It does not, however, state what assumption was actually used for the feasibility calculations. The discussion is further confused by the later paragraph that indicates a range of 4-fold lower to 4-fold higher; the report should probably illustrate circumstances in which the effective dose is twice the outside dose.
Page 55, third paragraph
“The actual dose to a person who lived in the United States during the years of fallout should generally lie within a range from one-fourth as large as these estimates provided here to about four times larger than these estimates.” How could such a conclusion be considered reliable on the basis of the limited study conducted?
Page 57, top of page
It is not clear how a person’s diet would affect estimates of external dose, which is what is discussed immediately before and after the statement. The statement made here is strictly correct in that it refers to estimates made throughout the report, but it is misleading in context.
Page 58, Table 3.4
It is not clear what the footnote in this table is to refer to or why it may be important.
The discussion of what was done for this feasibility study is misleading. Page 59 states that “the general ORERP method is described here because it was used for these feasibility calculations.” That is incorrect. Page 60 implies that various mathematical models were used as components of this feasibility study (or possibly as components of the ORERP method), but in fact published data that summarize the outputs of such mathematical models were used (even in ORERP). For example, the dose coefficients Fg are obtained from such mathematical models. However, the models were not themselves part of this feasibility study, as implied; the dose coefficients Fg were obtained from ICRP publications. Similarly, the integrated intakes, I, were not obtained by running the PATHWAY model, as implied in the discussion here. Instead, they were interpolated from some published outputs of the PATHWAY model. The discussion should be reformulated to indicate what was done.
Page 62, line 2
Change 91S to 91Sr.
Page 67, line 9
The text says that the thyroid doses to adults are generally lower by a factor of 10 than doses to a child born on January 1, 1951. However, Table 3.6 indicates that the differences in population-weighted values are only a factor of 6. Are the two statements correct and consistent?
Page 69, 2 lines below Table 3.7
Change “10 tests” to “16 tests”.
Page 81, lines 8–10
Clarify what is meant by “somewhat conservative”.
Page 86, line 5
Change “east” to “west”.
Table 3.14, Footnote a
The U.S. Census bureau (http://eire.census.gov/popest/archives/pre1980/popclockest.txt) indicates that the 1960 population of the United States was 1.81×108, not 1.63×108. Some explanation of the difference is desirable.
Page 102, Figure 3.24
The legend of Figure 3.24 and the Y-axis legend do not agree. It looks as though they should both refer to µSv PBq−1.
Page. 109, line 8 below Table 3.21
Delete “of” between “significant” and “amounts”.
Page 116, line 14
“Larger” than what?
Page 129, end of second paragraph
This sentence hints at, but does not say, that there may be no risk of heart disease and a variety of other nonneoplastic diseases at the doses under consideration. It should be explicitly stated.
Page 131, lines 1–2
It is an overstatement to say that the uranium-miner data constitute the only exception to the inferior contribution of radiation studies other than the atomic-bomb data. As one counterexample, the Canadian multiple-fluoroscopy study has more moderate- to high-dose data with respect to lung and breast cancer than the atomic-bomb study does.
Page 136, first full sentence
The committee suggests that the mechanistic rationale of Dilwyn Williams (2002) be mentioned as a reason for the age difference in radiosensitivity, namely, that the thyroid epithelial cells have the capacity for only a finite number of replications and that radiation exposure occurring when the cells are in replicative senescence in early adulthood will induce little or no cancer.
Page 138, last full sentence
The evidence that “the rate of increase in risk [of benign tumors] with dose is generally similar to that for cancer” is extremely small. One could make a case on the basis of various studies that benign-tumor risk is greater than, equal to, or less than the cancer risk, but the generality of any assertion would not be very convincing.
Page 139, middle of page
It is stated that “thyroid adenomas do not appear to have a malignant potential.” That is controversial; as some thyroid pathologists (such as Dilwyn Williams) believe the contrary, and a recent pooled study of thyroid-cancer etiology showed that thyroid-cancer risk is associated with a history of thyroid nodules (Franceschi et al. 1999).
Page 139, regarding benign thyroid tumors
“There are no data that provide quantitative information on risk estimates” is an overstatement. One can derive quantitative information (albeit of variable quality) from a variety of studies, such as those on tinea capitis (Ron et al., 1989), the thymus (Shore et al., 1993), lymphoid hyperplasia (Pottern et al., 1990), thyrotoxicosis (Dobyns et al., 1974), NTS fallout (Rallison et al., 1990), high-background areas (Pillai et al., 1976; Wang et al., 1990), the Nagasaki atomic bomb (Nagataki et al., 1994), thymus and cervical adenitis (Maxon et al., 1980), and tonsils and nasopharynx (Royce et al., 1979).
Page 141, middle paragraph
“The excess non-neoplastic disease in the atomic-bomb survivors is seen at a level below that given for medical purposes” is a misstatement. It is not seen below medical diagnostic-procedure levels; “for medical purposes” should be changed to “for radiotherapy”.
Page 141, middle paragraph, last sentence
It is not just the “lack of a clear dose-response curve” that limits quantitative assertions but also the lack of support for an effect at low to moderate doses in animal studies. It would be more appropriate for the sentence to indicate that no effect has been demonstrated at low to moderate doses.
Page 142, end of middle paragraph
It is suggested that the last sentence read, “non-cancer health effects, and there is no clear evidence of any risk at low doses.”
Page 143, top of page
The discussion highlights the uncertainties in the risk projections and raises the question of whether, given the uncertainties, they “can be useful for developing public health policy.” The text then proposes to address the question by providing “preliminary example estimates of population risk.” In the case of radionuclides other than 131I, for which the incremental risk from fallout exposure is small, population risk provides little or no useful information for setting public-health policy, because “collective dose” does not convey the important fact that risk to the individual is almost negligible.
Page 144, end of first paragraph
This statement implies that the relationship between lifetime risk and radiation dose is linear. Earlier, it was stated that the leukemia risk was nonlinear. If linearity is to be assumed, the assumption should be explicitly stated and justified.
Page 145, first full paragraph
This discussion of using the “average risk” fails to note that the average risk represents the
average lifetime likelihood that people in a particular group will develop cancer and thus is itself more important than the “estimated number of radiation-induced cancers in the population.”
Page 146, third paragraph, last sentence
The statement that “statistical uncertainty tends to be larger when risks are small than when they are large” is true if the magnitude of uncertainty is viewed on a relative basis (the ratio of the uncertainty to the risk) but not on an absolute basis. Similarly, the statement beginning on page 148, four lines from the bottom, should be “The relative uncertainties…”
Page 146, bottom, and page 147, top
Other important potential biases in studies of radiation effects that are not mentioned are selection biases and surveillance biases.
Page 148, first paragraph, last sentence
That a threshold “hypothesis is not supported by currently available data” is an overstatement. Consider bone cancer, lymphoma, soft-tissue sarcoma, and possibly skin cancer. For each of those, the case for a threshold is moderately strong.
Page 149, next-to-last line
Why does the discussion of RBE refer to “tissue damage” rather than disease risk? Tissue damage has its own set of parameters that are not necessarily congruent with, for example, those of cancer induction.
Page 150, first paragraph
The text mentions the likelihood that relative risk decreases over time, but this is not explored in the context of the risk estimates set forth to describe the possible effect of a temporal decrease in risk in reducing lifetime radiation risk. With regard to thyroid cancer, the most important cancer considered in this report, there is a highly significant reduction in thyroid-cancer relative risk with time; if factored in, it reduces the lifetime risk of thyroid cancer after childhood exposure by about 40–60% (Shore & Xue, 1999).
Page 150, four lines from bottom
To say merely that quantifying uncertainty “requires subjective judgments” makes it appear that the process is simply guessing. It might be better to indicate that judgments are based on the weight of available scientific evidence and attempt to characterize uncertainty on the basis of that evidence with bounds that reflect the available data.
Page 151, end of second paragraph
The sources of uncertainty in dose estimates should refer to the location in Chapter 3 where this information is given, and additional uncertainties in risk estimation (such as statistical variability, uncertainties about DDREF, and the shape of the dose-response curve) should be clearly summarized here, rather than requiring the reader to try to find them in Chapter 3.
Page 151, last full sentence
This sentence is ambiguous. Is it saying that only NTS thyroid doses would be received in childhood, so that the global thyroid doses that would be received mainly a decade or more later
would not have much effect? (That would be true only of those who were born in the early 1950s, not of other birth cohorts.) Or is it saying that thyroid doses from global fallout were much lower than those from NTS fallout? The sentence needs to be reworded to be clear and accurate.
Page 153, bottom, and page 154, top
The sentence juxtaposes the estimated 11,000 lifetime cancer deaths from fallout with the 500,000 yearly “spontaneous” cancer deaths—an inappropriate comparison because one is yearly and the other is lifetime. Although it next mentions the lifetime “spontaneous” cancer deaths, the text is misleading. The next sentence uses a figure of 22,000 excess cancer cases but gives nothing to compare that number with.
Page 154, middle paragraph, line 6
It should read “(a lifetime risk of about 1 in 16,500)”.
Page 155, Table 4.1
This table should include a column to give the percentage of cancers that are estimated to be caused by fallout (or perhaps more understandable wording would be, for example, “one of every 3,300 cancers was possibly caused by fallout”). The table (or a separate table) should also present a similar summary for the total population of 163–250 million who are described in the text.
There seems to be a discrepancy: the top of the page gives a median estimate of 7,400 cancer deaths, but the last paragraph says 11,000.
Page 157, line 2
That the total-body doses were low and had relatively little variation by geographic region, age, and so on, “makes it more difficult to identify groups with particularly large risk” conveys the idea that the object of the exercise is to find groups with large risks. It might be better worded, “means that there are unlikely to be groups with large risks associated with fallout radiation”. Similar wording on page 166, last line, should also be changed; the present wording suggests that small regional variations in whole-body dose are unlikely to be correlated with regional variations in radiation-induced disease gradients.
Page 157, third paragraph
The statement that “increased rates of leukemia have been reported in persons living downwind of the Nevada Test Site (Stevens et al., 1990)” stretches the finding. The result was not statistically significant except in one subgroup. Similarly, an earlier study of childhood leukemia among NTS “downwinders” found no excess leukemia (Lyon et al., 1979).
Pages 157–158, Section 220.127.116.11
The section does not mention the possibility of a threshold or a quadratic curve for leukemia, as shown in the atomic-bomb study.
Page 162, last paragraph
Differences in the natural frequency of disease—for example, high gastric cancer and low prostatic and breast cancer in the Japanese as contrasted with Caucasians pose a challenging problem. The statement that little can be done to reduce “uncertainties in using estimates from Japanese atomic bomb survivors for a United States population” appears to be premature. There has not yet been a thorough attempt to analyze the atomic-bomb study and the available studies in Western populations in parallel for a number of tumor sites so as to permit better extrapolations across populations. The current Biological Effects of Ionizing Radiation (BEIR VII) report may improve the state of affairs appreciably, although admittedly this is a difficult problem.
Page 163, first sentence
The claim that data were presented to describe the relationship between radiation and benign tumors of the thyroid, stomach, and other sites and heart disease is specious. No such data were presented.
Page 163, end of second paragraph
What is the basis for the speculation that “individuals with pre-existing disease could be more radiosensitive”? What pre-existing disease and what type of cancer are being referred to? Perhaps different genetic polymorphisms or mutations, and conditions such as ataxia telangiectasia and Bloom’s syndrome?
Page 165, middle paragraph
It would be appropriate to add the Hanford study results to the statement of no evidence of a radiation-related increase in nonneoplastic thyroid disease in Chernobyl.
Page 167, end of first paragraph
The text states that the “most likely non-cancer health outcomes are those affecting the thyroid gland” but does not state that other noncancer relationships are very unlikely.
Appendix 1 of Appendix D
This appendix presents an example related to exposure evaluation, using a method to evaluate cumulative exposure to external gamma radiation over the interval 20 d to 200 d relative to time of arrival (TOA). For gamma radiation from 137Cs deposited in soil, the exposure rate at time t during 20 to 200 days after TOA is given by
where Cs(20d), in mCi/km2, is the deposition density for 137Cs (per unit 131I) 20 d after detonation; the parameter value 6.15E-03 is in µR/h per mCi/km2; and the value 24 converts days to hours. While t is not explicitly defined, the equation and the value given of 6.15E-03 (Beck, 1980) requires that it be the time since 20d after TOA. The integral from 20 d to 200 d, I(mR), is indicated to be
Since t is time after 20 days from TOA, the second expression contains a systematic error: the cumulative exposure for the interval is underestimated by a factor of exp(20λ). If this is truly a calculation error, rather than simply a documentation error, it could be large. However, this appears not to be a problem for 137Cs, in that λ is small. The impact of the error would depend on how many of the short-lived radionuclides considered had substantial gamma emissions. External gamma-ray exposure-calculation methods used should be checked for the possibility of the indicated systematic error. If it is determined that such an error exists, its impact should be evaluated for each gamma-emitting radionuclide of interest.
The committee suggests that the glossary include radioactive elements referred to in the text with only chemical symbols and include explanations of various models (such as linear no-threshold and linear-quadratic). The models are important in interpreting data in Chapters 3 and 4.