ML20127F269

Forwards Critique of Aamodt Cancer Rept. Critique Based on Limited Data.Conclusions Reached by Aamodt Group Are Highly Untenable & Methodology Used Does Not Meet Even Most Lenient Epidemiologic Std

ML20127F269
Person / Time
Site:	Three Mile Island
Issue date:	08/29/1984
From:	Tokuhata G PENNSYLVANIA, COMMONWEALTH OF
To:	Hinson C Office of Nuclear Reactor Regulation
Shared Package
ML20127C068	List: ML20127C065 ML20127F223 ML20127F239 ML20127F252 ML20127F269 ML20127F296 ML20127F310
References
FOIA-85-8 NUDOCS 8506250097
Download: ML20127F269 (5)

Text

'

n a , m. . . .g 4 .d i- .

59t$& D{ ht11jyg

% '?

40 // f;7 Department of Health ,

7Uf"f5Y"((64 August 29, 1984 Mr. Charles Ilinson Radiological Assessment Branch U.S. Nuclear Regulatory Commission Room P 712 Washington, D.C. 20555

Dear Mr. Ilinson:

As you requested, we have enclosed our critique of the Aamodt Cancer Report which has been filed with the NRC Atomic Li-censing Board.

What is described in our critique is based on the materials made available to us at this time. With this limited data, there is ample evidence to indicate that conclusions reached by the Aamodt group are highly untenable and that methodology,they used does not meet even the most lenient epidemiologic standard.

In the event more source data were made available to us, we would be happy to provide another critical review for NRC.

Thank you for your interest.

Sincerely, 4

George K. kuhata, Dr.P.ll., Ph.D.

Director Division of Epidemiology Research GKT: ale Enclosure cc: Dr. Muller Dr. Reid Dr. Ilays Mr. Digon 8g6250]7850327 0

AANODTB5-8 PDR ~

P.o. BOX 90 H ARRISBURG. P A 17108 ,

• CRITIQUE OF 111E AAF10DT STUDY:

(Cancer Around Tbil)

Results of the Aamodt study were first reviewed on June 21, 1984, the day it was made public at a press conference in the Capitol Rotunda, Harrisburg, C Pennsylvania. At that time, Flarjoric and Norman Aamodt, intervenors in the Three blile Island (Till) Unit One restart case before the Nuclear Regulatory Commission, released a document which included results of the study. The doc-ument was titled, Aamodt blotions for Investigation of Licensee's Reports of Radioactive Releases During the Initial Days of the 'I111-2 Accident and Post-ponement of Restart Decision Pending Resolution of This Investigation.

According to information included in the document, the three areas "sel-

ected for inclusion in the survey were ones where residents had experienced erythema and metallic taste during the early days of the accident." One of the areas (Area 1) was six miles northwest of the plant and another (Area 2) three and one-half miles to the southwest. The third area (Area 3), seven miles northwest of the plant, was chosen because of its high elevation and clear view of the TF11 plants. The specific locations of the three areas were not provided.

The actual survey was conducted by a group of local resident women, some of whom p are reported to be experienced in conducting surveys (some of these interviewers h

are well recogni:cd anti-nuclear activists). The survey was stated to be organ-f' ized on the basis of information (advice and questionnaire form) provided by Dr. Carl Johnson of Denver, Colorado.

Fluch of the demographic cancer mortality data made available was included .

in Figure 2 - Cancer Death Rate Analysis. Without additional information, it was not possible to verify the data shown but in one instance, " Total Number of l flouseholds, 1979-1984" data were provided for only two of the three areas. In another, the method of computing the combined (three area) ratio of actual to l

expected deaths was incorrect. No information regarding cancer site or dates of diagnosis and/or death was provided. Information on other health effects included g

i numbers of cases of " spontaneously ruptured or collapsed organs", " persistent t I

rashes" and " birthing abnormalities." Year of occurrence was provided for each ,

s a of the four cases of " collapsed or ruptured organ" cases (collapsed lung,  !

. collapsed kidney, ruptured aortic valve) but the dates of occurrence of the -l

" birthing abnormalities" were not stated. Again the diagnostic information was vague and could not be verified. Much of the other health information is pro-vided in eight affidavits attached to the document.

m One statement in the document is particularly pu:: ling. "Several other residents of the TM1 area, not in the precise areas surveyed, but residing or working in the area northwest of the plants were also interviewed because we learned of their unique experiences." It is not stated if data obtained from

. these persons were included in the study. If they were included, the results

. would be seriously biased.

A review of availabic cancer mortality data from the State llealth Data Center for minor civil divisions in the Aamodt survey area did not indicate the existence of an apparent cancer problem but further evaluation of the Aamodt data was not

?

possible because detailed information was not available.

On August 15, 1984, additional data from the Aamodt study were made available l to the State !!calth Department through a member of TMI Public flealth Fund Advisory

)

Group, for Areas 1 and 2, the largest of the shey areas. These permitted a

! more comprehensive evaluation of the data included in the Aamodt document which  !

l cast serious doubt on the accuracy and utility of the study.

1 g A statement on Page 4 of the Aamodt document indicates that there were no w

[ refusals in Area 1 and four in Area 2. The summary data indicates that Case No.

b 162 refused information and that for Case No. 138 the number of people was unknown.

d j On the other hand, only two refusals could be found in Area 2 (Case No. 111, four i

l family members, and Case No. 207). The data in Figure 2 indicates that there were 40 households in Area 1 about which information was obtained but the maximum

-) appears to be 30. Similarly, 5b were reported for Area 2 but only 47 could be e

A -

-3 counted. With respect to the " Number of Persons About Information has Obtained" 112 were reported in Area 1, 88 were counted.

( More disturbing than the apparent differences in the above basic counts was the paucity of information about the cancer cases, as well as differences q in counts. Seven cancer deaths were reported for Area 1; only six could be counted including one who was diagnosed in 1978. There were nine deaths reported for Area 2 but only eight could be found on the summary sheets, including .one who was diagnosed in 1979. Even more disconcerting, however, is the fact that interviews were not conducted to btain information on five of the decedents.

The only information reported for these five was "Not interviewed - cancer - died."

A year of diagnosis was provideo for only three of the 16 reported cancer deaths in Areas 1 and 2. In most instances, even the month of death was not reported.

The accuracy / completeness of the diagnostic information, apparently not verified by medical records, is questionable. In several instances, it is not stated if the cancer patient is living or dead, hence comparison of counts can be tricky.

t Six (presumable) living cases were reported for Area 1; eight (including one h 1974 colostomy case) were counted. Ten living (?) cases were reported for Area t

2. Nine were counted but these include one each of the following: not interviewed,

! breast cysts, unspecified large tumor under arm, fibrous tumor, and breast cancer i

f (wife who lived in area prior to accicent). Other important information, such as length of residence, was inconsistently reported.

An effort to obtain additional clarifying information on the " spontaneously ruptered and collapsed organs" and the " birthing abnormalities" was nonproductive.

In summary, the quality and completeness of the Aamodt study data made avail-abicaresuchastocasiseriousdoubtonthevalidityofthereportedresults.

There are many unanswered questions about their methodology. (1) Exact geographic

l. locations of people included in the study? (2) If everyone in all these areas was included (denominator)? (3) Accuracy of cancer diagnosis regarding organ site and date of diagnosis (numerator)? (4) Number of cancer cases already dead

i and number still living? (5) Accurate information on radiation exposure for each included in the study (for both who developed cancer and who did not develop cancer)? Without these essential data, neither mortality nor incidence rates  !

can be computed. -

I l

l George K. okuhata , Dr.P.H.. , Ph.D. {

Director 3 Division of Epidemiology Research 1 1

l l

l i

l

)

i i

l i

l i

l I

l

,