A comprehensive dose reconstruction methodology for former Rocketdyne/Atomics International radiation workers

Article (PDF Available)inHealth Physics 90(5):409-30 · June 2006with62 Reads
DOI: 10.1097/01.HP.0000183763.02247.7e · Source: PubMed
Abstract
Incomplete radiation exposure histories, inadequate treatment of internally deposited radionuclides, and failure to account for neutron exposures can be important uncertainties in epidemiologic studies of radiation workers. Organ-specific doses from lifetime occupational exposures and radionuclide intakes were estimated for an epidemiologic study of 5,801 Rocketdyne/Atomics International (AI) radiation workers engaged in nuclear technologies between 1948 and 1999. The entire workforce of 46,970 Rocketdyne/AI employees was identified from 35,042 Kardex work histories cards, 26,136 electronic personnel listings, and 14,189 radiation folders containing individual exposure histories. To obtain prior and subsequent occupational exposure information, the roster of all workers was matched against nationwide dosimetry files from the Department of Energy, the Nuclear Regulatory Commission, the Landauer dosimetry company, the U.S. Army, and the U.S. Air Force. Dosimetry files of other worker studies were also accessed. Computation of organ doses from radionuclide intakes was complicated by the diversity of bioassay data collected over a 40-y period (urine and fecal samples, lung counts, whole-body counts, nasal smears, and wound and incident reports) and the variety of radionuclides with documented intake including isotopes of uranium, plutonium, americium, calcium, cesium, cerium, zirconium, thorium, polonium, promethium, iodine, zinc, strontium, and hydrogen (tritium). Over 30,000 individual bioassay measurements, recorded on 11 different bioassay forms, were abstracted. The bioassay data were evaluated using ICRP biokinetic models recommended in current or upcoming ICRP documents (modified for one inhaled material to reflect site-specific information) to estimate annual doses for 16 organs or tissues taking into account time of exposure, type of radionuclide, and excretion patterns. Detailed internal exposure scenarios were developed and annual internal doses were derived on a case-by-case basis for workers with committed equivalent doses indicated by screening criteria to be greater than 10 mSv to the organ with the highest internal dose. Overall, 5,801 workers were monitored for radiation at Rocketdyne/AI: 5,743 for external exposure and 2,232 for internal intakes of radionuclides; 41,169 workers were not monitored for radiation. The mean cumulative external dose based on Rocketdyne/AI records alone was 10.0 mSv, and the dose distribution was highly skewed with most workers experiencing low cumulative doses and only a few with high doses (maximum 500 mSv). Only 45 workers received greater than 200 mSv while employed at Rocketdyne/AI. However, nearly 32% (or 1,833) of the Rocketdyne/AI workers had been monitored for radiation at other nuclear facilities and incorporation of these doses increased the mean dose to 13.5 mSv (maximum 1,005 mSv) and the number of workers with >200 mSv to 69. For a small number of workers (n=292), lung doses from internal radionuclide intakes were relatively high (mean 106 mSv; maximum 3,560 mSv) and increased the overall population mean dose to 19.0 mSv and the number of workers with lung dose>200 mSv to 109. Nearly 10% of the radiation workers (584) were monitored for neutron exposures (mean 1.2 mSv) at Rocketdyne/AI, and another 2% were monitored for neutron exposures elsewhere. Interestingly, 1,477 workers not monitored for radiation at Rocketdyne/AI (3.6%) were found to have worn dosimeters at other nuclear facilities (mean external dose of 2.6 mSv, maximum 188 mSv). Without considering all sources of occupational exposure, an incorrect characterization of worker exposure would have occurred with the potential to bias epidemiologic results. For these pioneering workers in the nuclear industry, 26.5% of their total occupational dose (collective dose) was received at other facilities both prior to and after employment at Rocketdyne/AI. In addition, a small number of workers monitored for internal radionuclides contributed disproportionately to the number of workers with high lung doses. Although nearly 12% of radiation workers had been monitored for neutron exposures during their career, the cumulative dose levels were small in comparison with other external and internal exposure. Risk estimates based on nuclear worker data must be interpreted cautiously if internally deposited radionuclides and occupational doses received elsewhere are not considered.
Paper
A COMPREHENSIVE DOSE RECONSTRUCTION
METHODOLOGY FOR FORMER ROCKETDYNE/ATOMICS
INTERNATIONAL RADIATION WORKERS
John D. Boice, Jr.,*
Richard W. Leggett,
Elizabeth Dupree Ellis,
§
Phillip W.
Wallace,
§
Michael Mumma,* Sarah S. Cohen,* A. Bertrand Brill,
Bandana Chadda,*
Bruce B. Boecker,** R. Craig Yoder,
††
and Keith F. Eckerman
Abstract—Incomplete radiation exposure histories, inadequate
treatment of internally deposited radionuclides, and failure to
account for neutron exposures can be important uncertainties
in epidemiologic studies of radiation workers. Organ-specific
doses from lifetime occupational exposures and radionuclide
intakes were estimated for an epidemiologic study of 5,801
Rocketdyne/Atomics International (AI) radiation workers en-
gaged in nuclear technologies between 1948 and 1999. The
entire workforce of 46,970 Rocketdyne/AI employees was
identified from 35,042 Kardex work histories cards, 26,136
electronic personnel listings, and 14,189 radiation folders
containing individual exposure histories. To obtain prior and
subsequent occupational exposure information, the roster of
all workers was matched against nationwide dosimetry files
from the Department of Energy, the Nuclear Regulatory
Commission, the Landauer dosimetry company, the U.S.
Army, and the U.S. Air Force. Dosimetry files of other worker
studies were also accessed. Computation of organ doses from
radionuclide intakes was complicated by the diversity of
bioassay data collected over a 40-y period (urine and fecal
samples, lung counts, whole-body counts, nasal smears, and
wound and incident reports) and the variety of radionuclides
with documented intake including isotopes of uranium, pluto-
nium, americium, calcium, cesium, cerium, zirconium, tho-
rium, polonium, promethium, iodine, zinc, strontium, and
hydrogen (tritium). Over 30,000 individual bioassay measure-
ments, recorded on 11 different bioassay forms, were ab-
stracted. The bioassay data were evaluated using ICRP bioki-
netic models recommended in current or upcoming ICRP
documents (modified for one inhaled material to reflect site-
specific information) to estimate annual doses for 16 organs or
tissues taking into account time of exposure, type of radionu-
clide, and excretion patterns. Detailed internal exposure sce-
narios were developed and annual internal doses were derived
on a case-by-case basis for workers with committed equivalent
doses indicated by screening criteria to be greater than 10 mSv
to the organ with the highest internal dose. Overall, 5,801
workers were monitored for radiation at Rocketdyne/AI: 5,743
for external exposure and 2,232 for internal intakes of radio-
nuclides; 41,169 workers were not monitored for radiation.
The mean cumulative external dose based on Rocketdyne/AI
records alone was 10.0 mSv, and the dose distribution was
highly skewed with most workers experiencing low cumulative
doses and only a few with high doses (maximum 500 mSv).
Only 45 workers received greater than 200 mSv while em-
ployed at Rocketdyne/AI. However, nearly 32% (or 1,833) of
the Rocketdyne/AI workers had been monitored for radiation
at other nuclear facilities and incorporation of these doses
increased the mean dose to 13.5 mSv (maximum 1,005 mSv)
and the number of workers with >200 mSv to 69. For a small
number of workers (n 292), lung doses from internal
radionuclide intakes were relatively high (mean 106 mSv;
maximum 3,560 mSv) and increased the overall population
mean dose to 19.0 mSv and the number of workers with lung
dose >200 mSv to 109. Nearly 10% of the radiation workers
(584) were monitored for neutron exposures (mean 1.2 mSv) at
Rocketdyne/AI, and another 2% were monitored for neutron
exposures elsewhere. Interestingly, 1,477 workers not moni-
tored for radiation at Rocketdyne/AI (3.6%) were found to
have worn dosimeters at other nuclear facilities (mean external
dose of 2.6 mSv, maximum 188 mSv). Without considering all
sources of occupational exposure, an incorrect characteriza-
tion of worker exposure would have occurred with the poten-
tial to bias epidemiologic results. For these pioneering workers
in the nuclear industry, 26.5% of their total occupational dose
(collective dose) was received at other facilities both prior to
and after employment at Rocketdyne/AI. In addition, a small
number of workers monitored for internal radionuclides con-
tributed disproportionately to the number of workers with
high lung doses. Although nearly 12% of radiation workers
had been monitored for neutron exposures during their career,
the cumulative dose levels were small in comparison with other
external and internal exposure. Risk estimates based on nu-
clear worker data must be interpreted cautiously if internally
deposited radionuclides and occupational doses received else-
where are not considered.
Health Phys. 90(5):409 430; 2006
Key words: dose assessment; exposure, occupational; epidemi-
ology; nuclear workers
* International Epidemiology Institute, 1455 Research Blvd.,
Suite 550, Rockville, MD 20850;
Vanderbilt University Medical
School and Vanderbilt-Ingram Cancer Center, Nashville, TN;
Oak
Ridge National Laboratory, Oak Ridge, TN;
§
Oak Ridge Associated
Universities, Oak Ridge, TN; ** Lovelace Respiratory Research
Institute, Albuquerque, NM;
††
Landauer, Inc., Glenwood, IL.
For correspondence or reprints contact: John D. Boice, Jr.,
International Epidemiology Institute, 1455 Research Blvd., Suite 550,
Rockville, MD 20850, or email at john.boice@vanderbilt.edu.
(Manuscript received 18 October 2004; revised manuscript re-
ceived 13 July 2005, accepted 11 November 2005)
0017-9078/06/0
Copyright © 2006 Health Physics Society
409
INTRODUCTION
E
PIDEMIOLOGIC STUDIES of radiation workers are conducted
to gain direct knowledge of low dose and low dose-rate
exposures (Gilbert et al. 1993a and b; Frome et al. 1997;
Cardis et al. 1995, 2005; Muirhead et al. 1999; Omar et
al. 1999; Gilbert 2001; Sont et al. 2001; Iwasaki et al.
2003; Boice 2006). Such studies can also validate current
radiation risk estimates based on higher doses delivered
at higher dose rates. Inherent limitations of occupational
studies include relatively small numbers of workers at
individual facilities and relatively low cumulative doses
even when several studies are combined (Gilbert et al.
1993b; Cardis et al. 1995, 2005; UNSCEAR 2000).
Sample size and exposure levels affect statistical power,
i.e., the ability of a study to detect an effect given that
there is one. However, there are other sources of uncer-
tainty that can distort study findings. Although personnel
monitoring devices and bioassay measurements provide
estimates of radiation dose in ways far superior than
possible for chemical or other agents found in the
workplace, they are nonetheless subject to random error
and to systematic biases (Gilbert and Fix 1995; Gilbert et
al. 1996; Gilbert 1998; Daniels et al. 2005; Daniels and
Schubauer-Berigan 2005). Random error in the measure-
ment of dose is independent across workers and tends to
reduce the power for detecting effects. Systematic error
or bias can lead to spurious results. Systematic biases can
include inadequate collection of prior or subsequent
radiation work histories, inadequate treatment of the
intake of radionuclides, and underestimation of neutron
exposures. Other sources of possible bias include medi-
cal radiation procedures, natural background radiation,
and even conventions in recording radiation dose, e.g.,
for doses below a minimum detectable level (MDL) of
the measurement device setting the dose to either zero or
to the MDL, or assigning a value for missing dose as the
maximum allowed by regulation during the reporting
period. Measurement uncertainties include those associ-
ated with differences in photon energy, exposure geom-
etry, and type of dosimeter.
In this paper we address the magnitude of several
sources of systematic bias within the context of a
comprehensive dose reconstruction study of Rocketdyne/
Atomics International (AI) workers (Fig. 1). Three
sources of potential bias will be evaluated: radiation
exposures received elsewhere, internal intakes of radio-
nuclides, and neutron exposures. Inadequate or improper
treatment of these exposures would tend to underestimate
organ doses and overestimate derived risk per unit dose.
This dose reconstruction program was conducted in
support of an epidemiologic study to identify health
effects on Rocketdyne/AI workers exposed to radiation
(Boice et al.)
‡‡
and not for support of compensability
decisions (Office of Workers’ Compensation Programs
2005; NRC 2003, 2005). The paper does not focus on all
sources of uncertainty but on the three mentioned for
which serious bias could result if not handled properly in
the context of the current study.
Since 1996, Rocketdyne had been owned by The
Boeing Company and was sold in 2005 to Pratt &
Whitney. Previous corporate owners were Rockwell
International (1973–1996), North American Rockwell
(1967–1973), and North American Aviation (1928
1967). North American Aviation established Rocketdyne
as a separate division in 1955, and Atomics International
was also established as a division that same year.
Atomics International merged with Rocketdyne in 1984.
Throughout this paper, “radiation workers at Rocket-
dyne” is meant to include all radiation workers at the
Santa Susana Field Laboratory (SSFL) and nearby facil-
ities in California regardless of corporate ownership at
the time of occupational exposure to radiation.
Between 1948 and 1999, thousands of Rocketdyne/
Atomics International workers were involved in a wide
‡‡
Boice JD, Jr, Cohen SS, Mumma MT, Ellis ED, Eckerman KF,
Leggett RW, Boecker BB, Brill AB, Henderson BE. Mortality among
radiation workers at Rocketdyne (Atomics International), 1948 –1999.
Fig. 1. Flowchart of procedure for radiation dose determination for
Rocketdyne/AI workers.
410 Health Physics May 2006, Volume 90, Number 5
range of activities such as sodium-cooled breeder reactor
technology, uranium fuel fabrication, spent fuel evalua-
tion, radiography, hot lab chemistry, plutonium fuel
fabrication and storage of nuclear material. During the 52
years covered by the study, 5,801 workers were moni-
tored for external or internal radiation: 3,569 external
only, 58 internal only, and 2,174 both internal and
external (Table 1). Bioassay measurements of radioac-
tivity in urine and feces and whole-body and lung counts
were recorded for workers. Monitored radionuclides
included isotopes of uranium, plutonium, americium,
thorium, polonium, cesium, cerium, calcium, iodine,
zinc, zirconium, promethium, strontium, and hydrogen
(tritium).
A number of workers were employed at other
nuclear facilities before being hired by Rocketdyne/AI,
and many also were subsequently employed in radiation
occupations after leaving Rocketdyne/AI. To better char-
acterize the lifetime dose received by the workforce from
all sources, attempts were made to obtain additional data
through record linkage with dosimetry files available
from the Department of Energy, the Nuclear Regulatory
Commission, the Landauer dosimetry company, the U.S.
military services, and nine individual nuclear facilities or
prior study databases. To be included in this study,
workers had to have been monitored for radiation at the
SSFL or nearby facilities in California. The study re-
ceived human subjects research approval from Vander-
bilt University, The Boeing Company, and the Oak
Ridge Site-Wide Institutional Review Boards.
METHODS
The approach for obtaining lifetime career doses and
internal radiation doses to specific organs involved data
imaging, abstraction of internal monitoring documents,
obtaining external radiation exposure histories at Rock-
etdyne/AI, record linkage with nationwide and facility-
specific dosimetry files for external dosimetry outside of
Rocketdyne/AI, and internal dosimetry and biokinetic
modeling.
Data imaging
The goal of the imaging and abstraction process was
to create an electronic file for use in an organ dose
reconstruction scheme. Internal and external ionizing
radiation monitoring data from hard-copy records for the
employees at SSFL and other nearby Rocketdyne facili-
ties were evaluated. Because of the wealth of information
available in the radiation worker folders, as well as the
Table 1. Demographic and job characteristics for eligible workers who were monitored for radiation at Rocketdyne (n
5,801).
Characteristic n % Characteristic n %
Sex Male 5,335 92.0 Years of employment 0.5−1 215 3.7
Female 466 8.0 1−4 1,730 29.8
5−9 1,205 20.8
10−14 939 16.2
Race White 4,695 80.9 15−19 579 10.0
Nonwhite 340 5.9 20 748 12.9
Missing 766 13.2 Unk/Miss/Incorrect 385 6.6
Year of birth 1920 937 16.2 Radiation monitoring at External only 3,569 61.5
1920−1929 1,670 28.8 Rocketdyne External and internal 2,174 37.5
1930−1939 1,701 29.3 Internal only 58 1.0
1940−1949 769 13.3
1950−1959 534 9.2 Sources of radiation Rocketdyne 5,801 100.0
1960 190 3.3 dosimetry information Landauer 1,333 23.0
Dept of Energy 1,044 18.0
Nuclear Reg Comm 1,038 17.9
Year of hire 1948 98 1.7 U.S. Air Force 34 0.6
1948−1959 2,471 42.6 U.S. Army 57 1.0
1960−1969 1,963 33.8 U.S. Navy 26 0.4
1970−1979 607 10.5 Individual sources 64 1.1
1980−1989 595 10.3
1990 67 1.2 Number of sources 1 (Rocketdyne only) 3,212 55.4
providing radiation 2 1,713 29.5
dosimetry information 3 779 13.4
4 89 1.5
Year of 1960 319 5.5 5 8 0.1
termination 1960−1969 2,370 40.9
1970−1979 924 15.9 Vital status as of 12/31/1999 Alive 4,186 72.2
1980−1989 844 14.5 Dead 1,468 25.3
1990−1999 817 14.1 Died outside US 5 0.1
Active (12/31/1999) 527 9.1 Lost to follow-up 142 2.4
411Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
complexity of the bioassay report forms used over the
years, it was decided to scan and image all the folder
information into a searchable database. Among the
54,384 employees initially identified, 14,189 were po-
tential radiation workers based on having a radiation
folder.
Administrative practices at the time required that
each employee be issued a “radiation folder,” whether or
not he or she would ever be monitored for radiation or
actually receive occupational exposure to radiation. The
folders for every potential radiation worker were scanned
and indexed by name, social security number, worker
serial numbers, date of birth, and date of hire. There were
7,204 workers subsequently excluded from the cohort
when it was established because they were not occupa-
tionally exposed to radiation. These workers were ex-
cluded because they were never monitored for radiation
exposure and had no dosimetry information in their
folders. There were also 350 workers excluded because
they worked for less than 6 months, and 524 workers
because of insufficient identifying information. All of the
scanned images of the radiation monitoring records were
prescreened to identify those data folders that contained
bioassay data.
Abstraction of internal monitoring documents
The work proceeded as follows: documentation of
data flow and procedures, prescreening the scanned
images, data entry, quality assurance, and quality control.
Documentation of data flow and procedures. All
activities associated with prescreening of the imaged
files and construction of the electronic file containing the
internal radiation monitoring data from the scanned
images of the hard-copy records were documented.
Prescreening the scanned images. The radiation
monitoring records were scanned in batches according to
the likelihood of radiation exposure based on length and
calendar year of employment, facility of employment,
and information available from the prior investigation
(Ritz et al. 1999). Workers most likely to have the
highest exposure to radionuclide intakes were scanned
first to assure that proper effort was made to reconstruct
these rather complex dosimetry situations. As a result of
the initial records review, 11 different types of records
containing bioassay data were identified. Fig. 2 is an
example of a bioassay data form used after 1962.
Because of the number and variations in the layout
of the documents to be entered, speed and accuracy in the
entry process were facilitated by grouping like docu-
ments together. A folder was created on the desktop for
each of the 11 types of internal radiation monitoring
records. Each scanned image in the worker’s radiation
folder was then evaluated, and a copy of the image was
placed in the appropriate folder. A unique number
assigned to each image as it was scanned was used to link
the record to the correct worker. The few documents of
interest with a different format from the 11 common
Fig. 2. Master record for internal monitoring data, starting about 1963. Date is day of sample collection. In this example,
types of measurements were urinalyses for uranium (radiometric UR, code 1B; fluoroscopic UF, code 1A) and
mixed fission products (MFP, code 2B), and total-body counting of
235
U (TBC). For a urine sample collected on 21 Aug
67, for example, results for UR, UF, and MFP are 58.4 (dpm d
1
), 34.6 (
gd
1
), and 0 (dpm d
1
), respectively, assuming
a daily urine volume of 1,500 mL. Some cards in the 1950’s and early 1960’s used a /0” notation system to indicate
positive and negative bioassays. Numeric values were then provided on bioassay documents in the folder.
412 Health Physics May 2006, Volume 90, Number 5
document types were placed in a separate folder to be
data entered by the supervisor of the data entry clerks.
Data entry. Data entry was accomplished using a
dual monitor approach with one monitor used for view-
ing the scanned image and the second monitor used for
inputting the data. The result was a paperless process
designed to increase productivity and reduce data entry
cost. MS Access (Microsoft Corporation, One Microsoft
Way, Redmond, WA 98052-6399) was used for data
entry. The folders of images created during the pre-
screening were used to access the records pertaining to
the internal radiation monitoring of workers. A data entry
form was developed for each document type. Because of
the variety of document types and the number of vendors
used, all documents in a given folder (and of a given
type) were entered together. Over 30,000 internal radia-
tion monitoring records were processed.
Quality assurance. Quality assurance (QA) proce-
dures were established to confirm the accuracy of the
data file created and to promote cost efficiency and
timeliness. These procedures verified both the complete-
ness of the prescreening of the records and the accuracy
of the data entry. A sampling plan was developed to
specify sample size and associated acceptance criteria
based on statistical standards. The basic sampling unit
was an entire record that was evaluated as either defec-
tive or not defective. Clear operational definitions were
developed to insure that errors were assessed consistently
and precisely.
When data entry on a new form was started, it was
closely monitored to assure that the data entry clerks
understood clearly the instructions for entry and had
opportunities to ask questions. There was also real time
clean up on mismatched records, i.e., when the record in
the scanned file was incorrectly assigned to a specific
worker. For example, when a scanned file was found that
contained records for more than one person or the wrong
person, the linkage was changed to the correct worker.
These mismatches were identified by both the data
administrator and the data entry clerk.
Quality controls. After data entry and verification,
quality control (QC) activities were carried out on the
second electronic file created from the initial electronic
file. This second file was used to perform several edit
checks. The first compared the range of dates in the
monitoring data against the hire and termination dates for
the worker available from Kardex work histories or
electronic personnel listings. The second checked the
completeness of the individual numeric results from the
bioassay sample against the /0” values recorded on
the summary cards for the time period 1 January 1961
through 31 December 1967. The /0” designators on
the summary cards indicated the availability of numeric
bioassay results recorded on separate documents in the
worker’s radiation folder and were only used for the
period 1961–1967. After that, the numeric values were
recorded on the summary card.
Results of processing internal monitoring docu-
ments. There were 952 workers included in the internal
dosimetry initial selection, i.e., those who likely received
the highest intakes of radionuclides based on job loca-
tion, calendar year, and available Rocketdyne bioassay
data. An MS Access database containing a total of 27,023
records of internal radiation monitoring data was com-
piled from the scanned images of interest for these
workers. A copy of this file was provided to the internal
dosimetrists for use in reconstructing doses for these
workers.
In addition, we visually reviewed over 200,000
records within the 14,189 worker folders for any indica-
tion of either internal or external dosimetry. Subse-
quently, 1,280 additional workers were identified as
monitored for possible intakes of radionuclides, although
the bioassay data indicated no more than minimal levels
of detection for the assay employed. Overall, 2,232
workers were determined to have worked in areas with
the potential for radionuclide intakes and were so mon-
itored.
Finally, to supplement the information available in
the scanned folders of radiation workers, we also evalu-
ated any incident reports of potential unusual exposures
or wounds occurring in a controlled area as well as
dosimetry workups made at an individual’s personal
request. Complete dosimetry assessments had been per-
formed on 410 workers by the Rocketdyne health physics
staff based on worker request or for compensation
considerations.
Obtaining external radiation exposure histories at
Rocketdyne/AI
The overall plan was to obtain a complete career
history of radiation exposure for each worker, including
doses prior, during, and after employment at Rocket-
dyne/AI. All 14,189 Rocketdyne/AI radiation folders
were reviewed, and those with evidence of radiation
monitoring were selected. Because of an administrative
policy to issue a radiation folder to all new employees
regardless of their potential for radiation exposure mon-
itoring, 7,204 workers were excluded as not being
radiation workers. These excluded workers had never
worked in an area requiring a radiation-monitoring de-
vice and they had no dosimetry information in their
413Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
folders. Kardex job histories were sought to identify any
workers without monitoring information, and discussions
with former employees were held to confirm that there
was no group of non-monitored employees who consis-
tently worked in radiation areas. Similar to what was
done for the internal radiation exposure, external radia-
tion records were selected and sorted by type of form/
vendor to facilitate data entry. Entry into both MS Excel
and MS Access databases included name, social security
number, calendar year of exposure, photon radiation dose
for that calendar year, neutron dose for that calendar
year, and an indicator that the dose was received prior to
Rocketdyne employment, during employment (whether
onsite or offsite), or subsequent to Rocketdyne employ-
ment. All dosimetry data after 1990 had been computer-
ized and were accessible directly. There were 5,743
workers identified as having been monitored for external
radiation at Rocketdyne facilities.
Care was taken to exclude individuals who were
working at Rocketdyne as contractors. Contract workers
could be identified from a computerized database of
4,675 contract workers. The contractors were not Rock-
etdyne employees, did not have work histories, did not
have worker serial numbers, and did not have complete
dosimetry information. Because of the various exclusion
criteria and the lack of identifying information for many
workers, it was not possible to determine the actual
number of contract workers excluded. At least several
hundred contract workers were excluded, but the actual
number could have been much more.
Record linkage with nationwide and facility-specific
dosimetry files for external radiation outside of
Rocketdyne
Anecdotal information combined with evidence in a
number of dosimetry files suggested that a significant
number of Rocketdyne/AI workers had been exposed to
ionizing radiation at other facilities, both before and after
working at Rocketdyne. For an accurate assessment of
the risk from occupational radiation, it was essential that
complete lifetime doses be determined. External radia-
tion exposure records were obtained from a variety of
sources in an attempt to characterize lifetime occupa-
tional exposures for all workers. The dosimetry data
found in the Rocketdyne/AI radiation folders, which
formed the basis of the radiation worker database, were
then supplemented with dosimetry information obtained
from the Landauer dosimetry company, the Nuclear
Regulatory Commission, the Department of Energy, and
the military services (e.g., the U.S. Army) (Muirhead et
al. 1996). In addition, nuclear facilities where a worker
was employed prior to joining Rocketdyne or transferred
to after employment with Rocketdyne/AI were identified
and dosimetry information obtained (Gilbert et al. 1993a;
Fry et al. 1996; Frome et al. 1997; Dupree-Ellis et al.
2000).
Permissions were sought to access various national
dosimetry databases using name, social security number,
and date of birth as matching variables. The Nuclear
Regulatory Commission REIRS files (Radiation Expo-
sure Information and Reporting System) contained expo-
sure information by calendar year for workers who
terminated employment at any of the NRC licensee
facilities. The Department of Energy allowed access to
various databases including the REMS (Radiation Expo-
sure Monitoring System) files. The Landauer dosimetry
company has computerized records of their clients be-
ginning in 1977 and hard copy records back to 1953.
Landauer was the major source of additional dosimetry
from non-nuclear facilities, e.g., medical or industrial
radiography. The U.S. Army and U.S. Air Force had
computerized dosimetry records beginning as early as
1959 and 1962, respectively. Other sources of dosimetry
information included investigators of other worker stud-
ies who were asked to match their dosimetry files against
our study roster. In total, radiation dosimetry was ob-
tained from five major national databases, including 32
different nuclear facilities.
Matching the 5,801 Rocketdyne/AI radiation worker
files with the external dosimetry databases identified
1,833 workers (nearly 32% of all radiation workers) who
had been monitored for radiation exposure at other
facilities. Interestingly, matching the 41,169 Rocketdyne
non-radiation worker files with the same databases also
revealed that 1,477 (3.6%) of these workers had been
monitored for radiation doses at non-Rocketdyne facili-
ties.
Table 1 and Fig. 3 present the number of dosimetry
sources and number of workers for whom dosimetry
information was obtained. Over 36% of all radiation
workers were employed at facilities other than Rocket-
dyne, and some workers had been employed at as many
as five different facilities. By far, the Hanford site had
employed more Rocketdyne employees (1,194) than any
other installation, consistent with the fact that Rockwell
International was the third manager of the Hanford site
after Dupont and General Electric. Other frequent places
of employment included the Idaho National Engineering
Laboratory (INEL; 237), Rocky Flats (160), the Nevada
Test Site (103), Los Alamos National Laboratory (92),
the Oak Ridge site (95), Argonne National Laboratory
(74), the U.S. Air Force (152), and the U.S. Army (152).
Access to U.S. Navy dosimetry records was not obtained,
but information in the Rocketdyne radiation folders was
found for 26 workers who had been previously employed
in the Nuclear Navy (mean dose 8.0 mSv; range 0 45.3
414 Health Physics May 2006, Volume 90, Number 5
mSv). Throughout this manuscript we use the term
“dose” interchangeably with “equivalent dose” although
“equivalent dose” is technically correct when units of
mSv are used.
Dose information was available from seven overlap-
ping sources (Fig. 3, Table 2). Care was taken to assure
that only non-overlapping dosimetry information was
incorporated into the analyses. Occasionally, calendar
year exposures were not available, and for such instances
the cumulative dose, or termination dose, was recorded
in the calendar year in which it was reported. These
combinations occurred primarily for early doses obtained
prior to employment at Rocketdyne/Atomics Interna-
tional and thus will have little impact on the exposure
lagging analyses, i.e., even a 10-y lag period would be
unlikely to result in any of these exposures being
excluded. The major source of radiation exposure histo-
ries came from the Rocketdyne/AI files, which also
included documentation of prior radiation doses received
by over 400 workers. Other sources of information
included the Landauer dosimetry company (n 1,792),
which was the major vendor providing dosimetry ser-
vices for Rocketdyne/AI over the years, the Department
of Energy (n 2,058), which oversaw the Hanford site
and other national laboratories where Rocketdyne/Rock-
well International had management responsibilities, and
the Nuclear Regulatory Commission (n 1,039). Mean
doses are presented in Table 2 by source of radiation
history. The mean and collective doses for workers
monitored for radiation at Rocketdyne increased by 35%
when the doses received elsewhere were included, i.e.,
from 10.0 mSv to 13.5 mSv and from 57.4 person-Sv to
77.6 person-Sv. Of the 1,833 workers who had been
monitored for radiation other than at Rocketdyne/AI, 604
(or 10%) had greater occupational exposures elsewhere
than received at Rocketdyne. Also as noted, 1,477
workers who were not monitored for radiation at Rock-
etdyne were monitored elsewhere, although their mean
dose was low (2.6 mSv).
Neutrons. There were 584 radiation workers (10%)
who had been monitored for neutron exposures at Rock-
etdyne/AI and an additional 81 radiation workers who
were monitored for neutrons at other facilities: 617 had
5 mSv, 35 between 5–10 mSv, and 13 had between
10 60 mSv (mean 1.2 mSv; max 55.8 mSv). It was
presumed that quality factors of 10 for fast neutrons and
3 for slow neutrons had been used to reflect the relative
biological effectiveness of neutrons in comparison with
photons (ICRP 1991). These neutron doses were added to
the other external and internal doses received by each
worker. The contribution of neutron doses to total dose
for the 665 workers monitored for neutron exposures,
however, was small and less than 3% of the total external
photon dose they received. The contribution of neutron
dose to the collective dose from all external exposures
was accordingly small and only 0.8 person-Sv or 0.1%.
Examples of worker doses received elsewhere.
Just over 400 workers had a record of prior radiation
work that was found in the Rocketdyne/AI folders; over
500 additional workers were found through record link-
age to have had prior radiation exposure; and over 1,200
workers had been found through record linkage to have
left Rocketdyne and were monitored for radiation else-
where. Below are several examples where incomplete
knowledge of occupational radiation received elsewhere
would have led to serious underestimation of a worker’s
exposure.
One worker employed at Rocketdyne from 1961 to
1964 had received 42.9 mSv while so employed; he
had come from INEL with 131.3 mSv and received
10.8 mSv after leaving Rocketdyne and returning to
INEL for a total career dose of 185.0 mSv;
Another worker was employed at Rocketdyne/AI be-
tween 1962 and 1967 and received 11.2 mSv. He left
Rocketdyne for Hanford and received an additional
125.3 mSv for a total career dose of 136.5 mSv; and
Another worker was employed at Rocketdyne/AI be-
tween 1960 and 1961 and received 3.8 mSv total
radiation dose. After leaving, he worked at the Idaho
Fig. 3. Sources of radiation exposure histories.
415Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
Field Site and at Argonne National Laboratory where
he received an additional 157.1 mSv for a total career
dose of 160.9 mSv.
There were also 1,477 workers who were not mon-
itored for radiation at Rocketdyne/AI but were monitored
elsewhere. Several examples are given below.
One worker who began work at Rocketdyne in 1979
but was not monitored for radiation during his employ-
ment had received 186 mSv prior to Rocketdyne at
INEL; and
Another worker who left Rocketdyne in 1969 without
being monitored for radiation was subsequently mon-
itored at the Hanford site and received 188 mSv.
Internal dosimetry and biokinetic models
Assignment of doses from internally deposited
radionuclides. Annual doses from internally deposited
radionuclides were estimated in cases where intakes were
indicated by positive bioassay data, in vivo lung counts,
or incident reports. Screening criteria were developed to
reduce time-consuming analyses of relatively low intakes
unlikely to be meaningful with regard to the goals of the
epidemiological study. The screening criteria were de-
scribed in terms of total intake of specific radionuclides
but were based on the primary criterion that projected
lifetime equivalent doses from all intakes combined were
less than 10 mSv to any tissue. To implement this,
biokinetic models were used to develop intake levels
specific to different forms of radionuclides that produce
a 50-y equivalent dose of at least 10 mSv to the tissue
receiving the highest dose. Annual doses from internally
deposited radionuclides were calculated for those work-
ers (292) whose intakes met this criterion.
Internal monitoring data were found for 2,232 work-
ers. These consisted primarily of measurements of radio-
nuclides in urine, supplemented in many cases with fecal
measurements and external lung counts. For workers
assigned to areas with a relatively high potential for
internal exposure, urine samples generally were collected
at regular intervals. Sampling typically was on a quar-
terly basis but was more frequent if elevated exposure
was suspected. Follow-up measurements generally were
made if elevated internal exposure was detected.
In vivo lung counting was performed in many cases
involving suspected inhalation of enriched uranium and in
some cases involving suspected exposure to other radionu-
clides. Measurements of uranium in the lungs were gener-
ally reported as a mass of
235
U. Conversion to activity in the
lungs was based on information or assumptions regarding
the level of
235
U enrichment of the uranium to which the
worker was exposed. Over the years the uranium handled
by the Rocketdyne/AI workers varied in enrichment from a
few percent up to about 93%.
Intake estimates for workers sometimes were in-
ferred or adjusted on the basis of data for coworkers with
apparently similar exposures but more extensive moni-
toring data. For example, if nasal smears suggested that
two workers had similar intakes during an incident and
follow-up bioassay data were available for only one of
the workers, then those data were assumed to apply to the
other worker as well. Such use of surrogate data may
often overestimate intake because health physicists may
have based decisions regarding follow-up of individual
Table 2. Number of workers and mean external dose (mSv) by source of radiation history and whether or not workers
were monitored for radiation while at Rocketdyne.
Sources of dose information
Monitored for external radiation at Rocketdyne
a
Yes No Total
No.
Mean dose
(mSv) No.
Mean dose
(mSv) No.
Mean dose
(mSv)
Rocketdyne files 5,743 10.0 0 5,743 10.0
Landauer 1,333 21.7 459 2.7 1,792 16.9
DOE 1,042 8.3 1,016 2.7 2,058 5.5
NRC 1,037 18.0 2 3.0 1,039 17.9
U.S. Army 57 0.9 95 1.1 152 1.0
U.S. Air Force 34 1.9 118 0.3 152 0.7
U.S. Navy
b
26 8.0 0 26 8.0
Other sources 64 6.2 2 0 66 6.0
Total, unique workers 5,751
c
13.5 1,477 2.6 7,228 11.3
a
N 58 workers who are in the Rocketdyne Radiation Cohort but were only monitored for internal radiation are not included in this
tabulation.
b
Doses from the U.S. Navy are incomplete because no linkage was made of the entire cohort to Navy records. However, doses received
while in the Navy were obtained for some Rocketdyne workers from correspondence documents in the Rocketdyne radiation folders.
c
N 8 workers who were only monitored for internal radiation while at Rocketdyne were found to have received external radiation
prior to work at Rocketdyne and thus are included in this tabulation.
416 Health Physics May 2006, Volume 90, Number 5
workers on more information than appears in the work-
ers’ folders. On the other hand, use of surrogate data is
supported by a number of comparisons showing that
groups of workers involved in the same incidents often
showed similar urinary excretion of radionuclides over
an extended period.
Radionuclides addressed. Radionuclides listed in
the bioassay records of the Rocketdyne/AI workers include
isotopes of uranium, strontium, cesium, plutonium, ameri-
cium, zirconium, zinc, thorium, polonium, cerium, prome-
thium, calcium, iodine, and hydrogen (tritium). Those
radionuclides resulting in highest estimated internal doses at
this site are listed in Table 3. Reported data sometimes were
not specific to radionuclides but were given as total activity
of mixed fission products, gross alpha, gross beta, or other
non-specific terms. In such cases, the activity generally was
assigned to the dosimetrically dominant radionuclide
among those likely to be present in significant quantities.
The most frequent use of such a “surrogate radionuclide”
involved the assumption that activity reported as “mixed
fission products” consisted entirely of the long-lived bone
seeker
90
Sr. In most but not all such cases, this dosimetri
-
cally cautious assumption produced integrated dose esti-
mates lower than the screening level of 10 mSv described
earlier.
Models used to interpret bioassay data
Three general types of biokinetic models. Recon-
struction of doses from bioassay data was based on
biokinetic models that predict the time-dependent distri-
bution and excretion of radionuclides deposited in the
human body. The biokinetic models are of three main
types: a generic respiratory tract model that describes the
deposition and retention of inhaled material in the respi-
ratory tract and its subsequent clearance to blood or to the
GI tract; a generic GI tract model that describes the
movement of swallowed or endogenously secreted ma-
terial through the stomach and intestines and, together
with element-specific absorption fractions, its absorption
to blood; and element-specific systemic biokinetic mod-
els that describe the time-dependent distribution and
excretion of radionuclides after their absorption into
blood.
Respiratory tract model. Intakes were assumed to
be by inhalation unless there was reasonably good
evidence of intake through ingestion or a puncture
wound. The structure of the ICRP’s current respiratory
model (ICRP 1994a) was applied to all inhalation cases.
For each case, an “absorption type” was assigned to the
inhaled radionuclide on the basis of best available infor-
mation, such as records of the nature of the work and the
material being handled, the pattern of excretion of the
radionuclide over time, or the rate of decline of activity
in the lungs. Four absorption types for inhaled particu-
lates were considered: Type F, Type M, or Type S as
defined by the ICRP (ICRP 1994a, 1994b), or Type
“Modified S,” which was developed by the investigators
specifically for application to uranium aluminide.
Type F, Type M, and Type S, respectively, are
generic descriptions of rates and directions of transfer of
highly soluble, moderately soluble, and highly insoluble
material in the respiratory tract. They are referred to as
“absorption types” because the rate of absorption of an
inhaled radionuclide from the respiratory tract to blood
generally increases with the rate of dissolution of the
inhaled carrier in the tract.
Type F was applied to nearly all cases involving
inhalation of strontium, cesium, calcium, and unspecified
mixed fission products presumed to consist of
90
Sr or, in
some cases, a mixture of
90
Sr and
137
Cs. Application of
Type F to these cases was based mainly on recommen-
dations of the ICRP and, as illustrated in Fig. 4, is
generally supported by comparisons of model predictions
with excretion data in cases with detailed follow-up data.
Type M was used as the default absorption type for
isotopes of uranium, plutonium, thorium, cerium, prome-
thium, and polonium. Application of Type M to these
radionuclides is consistent with recommendations of the
ICRP and, as illustrated in Fig. 5, is generally supported
by comparisons of model predictions and urinary excre-
tion data in cases where workers exposed to these
radionuclides were followed for an extended period.
Special parameter values were developed for a form
of uranium called uranium aluminide that showed much
different initial behavior in the respiratory tract than
predicted by parameter values for Type M material.
Beginning in 1966, relatively high internal exposures to
uranium sometimes occurred in the “powder room” at the
De Soto facility, where large quantities of uranium
aluminide were handled. Bioassay data for powder-room
workers generally are not consistent with model predic-
tions based on any of the standard absorption types
addressed by the ICRP. This is illustrated in Fig. 6, where
urinary excretion data for a worker are compared with
model predictions for Type M or Type S material based
on air monitoring data. Curves for Type M or Type S
reflect the assumed intake pattern. In contrast to predic-
tions for these two absorption types, uranium aluminide
appeared to have a biphasic pattern of removal from the
lungs. As inferred from changes with time in the urinary
excretion rate, there was initially little dissolution of
inhaled particles, but the material presumably broke
417Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
down in the lungs over a few months and subsequently
was excreted at a rate consistent with moderately soluble
material.
This pattern of behavior is consistent with the
following theory based on observed properties of
uranium-aluminum intermetallic compounds prepared by
different techniques (Le Claire and Bear 1956; Buddery
et al. 1964; Giles and Tavender 1967; Bland 1968;
Subramanyam et al. 1985). Uranium is not distributed
completely uniformly in the preparation but may exist as
“islands” of unreacted uranium. These islands are seen
on microscopic examination as blisters that grow as the
underlying uranium is oxidized. The oxidation of ura-
nium islands may begin slowly but may accelerate as the
blisters grow, resulting in the breakdown of a uranium
aluminide particle. These descriptions are based on
microscopic examinations over dimensions approaching
those of respirable particles. Although the experiments
were not conducted under physiological conditions, it is
feasible that initially insoluble uranium aluminide parti-
cles could undergo the same gradual oxidation and
breakdown in the lungs.
The set of respiratory parameter values applied in
this study to uranium aluminide is a modification of the
ICRP’s parameter values for Type S material, i.e.,
relatively insoluble material with low absorption to blood
(Leggett et al. 2005). Parameter values describing reten-
tion of Type S material in the alveolar-interstitial region
of the lungs were modified to yield low absorption of
uranium to blood over the first few months after inhala-
tion. In effect, the modified Type S model depicts
uranium aluminide as an initially insoluble material with
very low absorption that is gradually transformed in the
lungs to a form that is more readily absorbed to blood.
The assumed rate of transformation from an insoluble to
a moderately soluble material is 0.004 d
1
, corresponding
to a half-time of roughly 0.5 y. With the exception of
transformed material in the alveolar-interstitial region,
particulate transport is the same as depicted in the ICRP’s
respiratory model for Type S material.
Gastrointestinal tract model. The model of the
gastrointestinal (GI) tract applied in this study is the
current GI model of the ICRP (ICRP 1979). The only
feature of that model of much importance to the present
study is the assumed GI absorption fraction for swal-
lowed activity. The GI absorption fractions or f
1
values
for inhaled radionuclides are those currently recom-
mended by the ICRP for workers (ICRP 1994b). The GI
absorption fraction for uranium inhaled as aluminide (not
addressed by the ICRP) is assumed to be 0.002, the
ICRP’s value for relatively insoluble forms of U.
Systemic biokinetic models. With three exceptions,
the systemic biokinetic models used to interpret bioassay
data and calculate organ doses are those currently rec-
ommended by the ICRP (ICRP 1993, 1994b, 1995a and
b, 1997). The ICRP’s current systemic biokinetic models
Table 3. Internally deposited radionuclides resulting in the highest internal dose estimates for Rocketdyne/AI workers.
a
Radionuclide Comments
234
U,
235
U,
238
U
Alpha emitters. Assumed to be moderately soluble in lungs except for U aluminide,
which is initially insoluble. Absorbed U mainly excreted in urine but some
deposition in bone, kidneys, and other tissues, with tenacious retention in bone.
239
Pu
Alpha emitter. Assumed to be moderately soluble in lungs when the compound is
not specified. Absorbed Pu divides mainly between bone and liver with 10%
distributed to other tissues. Extremely slow removal of absorbed Pu from body.
90
Sr
Usually relatively soluble in lungs. Absorbed Sr follows calcium and deposits
largely in bone, where a portion is retained for many years. For dosimetric
purposes,
90
Sr frequently was used as a surrogate for undetermined mixtures of
90
Sr,
137
Cs, and other fission products, i.e., for activity reported as mixed fission
products (MFP).
232
Th
Alpha emitter. Assumed to be moderately soluble in lungs when the compound is
not specified. Systemic biokinetics broadly similar to Pu but higher bone
deposition and lower liver deposition than Pu. Gives rise to chain of radionuclides
that migrate from parent and deliver much of the total dose.
210
Po
Alpha emitter. Usually moderately soluble in lungs. Absorbed Po concentrates
mainly in liver, kidneys, spleen, and bone marrow and is eliminated from body
over a period of months.
241
Am
Alpha emitter. Assumed to be moderately soluble in lungs. Biokinetics similar but
not identical to Pu; e.g., Am has slightly higher excretion rate than Pu.
144
Ce
Assumed to be moderately soluble in lungs. Biokinetics broadly similar to
239
Pu, but
dose per unit intake much lower for
144
Ce than
239
Pu due to shorter half-life of
144
Ce and lack of alpha emissions.
a
Radionuclides listed in approximate decreasing order of importance with regard to internal doses estimated for this site. Evidence of
internal exposure was also found for
137
Cs,
147
Pm,
131
I,
45
Ca,
3
H, and other radionuclides.
418 Health Physics May 2006, Volume 90, Number 5
for polonium, cerium, and promethium were replaced by
newer models provisionally adopted by the ICRP for use
in an upcoming document on occupational intakes of
radionuclides (Leggett and Eckerman 2001; Taylor and
Leggett 2003).
Application of models to estimate doses to work-
ers. Estimation of dose from excretion data or in vivo lung
counts is done in two steps: a “backward” calculation in
which the total intake is estimated on the basis of exposure
records and excretion or lung data, and a “forward” calcu-
lation in which annual doses are calculated on the basis of
the estimated intake. The backward calculation involves
determination of an appropriate exposure scenario (that is, a
characterization of the material taken into the body and the
pattern of exposure over time) and determination of an
intake level that provides a good fit between model predic-
tions and excretion and lung data based on the selected
exposure scenario, as illustrated in Figs. 4 and 5. In many of
the cases addressed in this study, the bioassay data changed
in a considerably less regular fashion than indicated in these
two figures. In such cases it was more practical, and
presumably no less accurate with regard to reconstruction of
dose, to fit cumulative excretion of radionuclides rather than
attempt to produce a model curve that mimicked the scatter
in the data. In such cases the fitting process consisted of
integrating urinary excretion data and then identifying a
level of intake that gives the same cumulative urinary
excretion value over the same period, based on the assigned
exposure scenario and biokinetic models.
Assignment of exposure scenarios. Estimates of
dose from internally deposited radionuclides may depend
strongly on the exposure scenario as well as the bioki-
netic models applied. Although exposure scenarios were
selected on a case-by-case basis and usually depended to
some extent on information specific to the exposed
worker, most of the cases represented some variation of
the general situations summarized below.
In a number of cases, the time of an acute intake of
a radionuclide was pinpointed in an incident report, and
the behavior of the internally deposited radionuclide over
time was revealed by follow-up measurements. Such
cases often allowed a check on the appropriateness of the
biokinetic models, as illustrated in Figs. 4 and 5.
Other cases of apparently short-term exposure in-
volve only portions of the information described above.
Fig. 4. Model predictions (curves) for acute intake of
90
Sr, Type F,
AMAD 5
m, compared with urinary excretion data for two
workers thought to have been acutely exposed to
90
Sr and other
fission products. Each curve is normalized to the observed excre-
tion rate 1 d after exposure.
Fig. 5. Model predictions (curves) for acute intake of uranium,
Type M, AMAD 5
m, compared with urinary excretion data
for two workers thought to have been acutely exposed to uranium.
Each curve is normalized to the observed excretion rate at 1 d.
Fig. 6. Urinary excretion data for a worker exposed to uranium
aluminide, compared with predictions based on reference absorp-
tion types used by the ICRP for moderately soluble (Type M) and
relatively insoluble (Type S) material. The intake scenario was
based on measured air concentrations during the exposure period.
419Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
For example, records indicate that a worker apparently
inhaled uranium oxide in early to mid-October 1965, but
the precise time of exposure is not known. Extensive
follow-up measurements are consistent with an elevated
intake of uranium around that time.
The time period of chronic exposure to a radionu-
clide often could be identified generally from collective
sources of information such as reports to the Atomic
Energy Commission, internal memos, air-monitoring
data, bioassay data, and individual work histories. For
example, the urinary data in Fig. 6 are for a worker
exposed to uranium aluminide from late 1966 to mid-
1967, according to reports and memos in his exposure
file. Air monitoring data indicating the change with time
in the concentration of uranium in air are available for
almost all of this period. A reasonable fit to this worker’s
urinary excretion data based on an exposure scenario
built from this information, together with the respiratory
model parameters developed for uranium aluminide, is
derived from an assumed intake of 100,000 Bq.
Common exposure scenarios were sometimes ap-
plied to groups of workers based on simultaneous
changes in their bioassay data. For example, a cluster of
positive urinary uranium measurements was found for
the first few weeks of 1963, with more than 50 workers
showing elevated urinary uranium during that period.
The patterns of change in the collective data over time
suggest that there may have been an incident during the
first few days of January 1963. The default assumption
for these workers was acute inhalation of moderately
soluble uranium (Type M) on 2 January 1963.
As an aid in assigning plausible scenarios to cases
with little direct information on the time-course of
exposure, a history of known or suspected internal
exposures to the Rocketdyne/AI workers was developed
in the form of a time line. This exposure time line was
based on incident reports and bioassay data extracted
from exposure histories of individual workers. Addition-
ally, interviews of former radiation workers were con-
ducted to learn first hand about incidents, exposure
circumstances, and work conditions. Approximately 200
potentially important radiation incidents involving a few
hundred workers were identified. Nearly all of these
cases involved acute releases of radionuclides, but the
term “incident” is used broadly here to refer to any
situation in which persons were potentially exposed to
elevated levels of one or more radionuclides, either
acutely or over an extended period.
Default methods applied when exposure patterns
were not evident. In numerous cases, intake of radionu-
clides was indicated by limited, irregular, or widely
spaced monitoring data, and no particular exposure
pattern was discernable from available information. In
such cases, default multipliers were used to estimate
intake. These multipliers were selected from distributions
of values derived from a variety of plausible exposure
scenarios. The multipliers are intended to be robust, i.e.,
to avoid large overestimates or underestimates in the
majority of cases. For example, an isolated urinary
uranium measurement of X Bq/d, with no indication of
the exposure time, was assumed to represent inhalation
of 1600X Bq of moderately soluble U (Type M). The
multiplier 1,600 was based on consideration of a variety
of plausible exposure scenarios, e.g., acute intake at
different times in the past six months, or different
patterns of chronic intake during that time. Most multi-
pliers derived from these scenarios were in the range
1,000 –5,000 and a cluster of estimates fell around 1,600.
Consideration of prior doses. Where feasible,
estimates of annual dose from internal exposure include
occupational intake of radionuclides prior to employment
at Rocketdyne/AI. Fig. 7 shows data for a worker whose
estimated internal exposures arose entirely from pre-
Rocketdyne/AI employment.
Treatment of less than minimal detectable levels
(<MDL). Assignment of a numerical value to a measure-
ment reported as MDL was determined on a case-by-case
basis. The value zero was assigned in the common situation
in which no recent exposure was suspected. If exposure was
suggested by ancillary information or if the MDL value
was close in time to positive measurements, then the value
0.5 MDL was usually assigned, but the value MDL was
assigned in a few cases where relatively high exposure was
suspected. The definition of “close in time” varied some-
what with the material under consideration. For example, it
was taken to be a few months for uranium aluminide, for
which urinary uranium may remain below the detectable
level for an extended period after elevated intake, and a few
weeks for moderately soluble uranium (Type M), for which
urinary uranium is tied more closely in time to the lung
burden.
Relative biological effectiveness of alphas. Radio-
biological data indicate that alpha particles and fission
neutrons have a larger biological effect than an equal
absorbed dose resulting from low-LET radiation. Ranges
of estimated values for the relative biological effective-
ness (RBE) of high-LET radiations are wide, depending
on the observed endpoint, the tissue, and perhaps the
animal species. Overall, experimental data for solid
tumor induction with alpha particles or fission neutrons
suggest a central value of about 10 –30 and a range of 6
420 Health Physics May 2006, Volume 90, Number 5
to 60 for the RBE relative to low-dose, low-LET radia-
tion (NCRP 1990; ICRP 1991; NRC-CEC 1997). The
RBE for leukemia appears to be considerably lower than
values reported for solid tumor induction (Boice 1993;
U.S. EPA 1999). In the present study, an RBE of 20 was
applied to absorbed dose from alpha radiation to tissues
other than red marrow (ICRP 1991), and an RBE of 1
was applied to absorbed dose to red marrow (Boice 1993;
U.S. EPA 1999).
RESULTS
Whole-body doses from external exposures
Whole-body doses from external exposures were
obtained in two ways: imaging and abstracting the
dosimetry information found within Rocketdyne/AI ra-
diation files and linking the roster of all Rocketdyne
workers with various nationwide databases and with
facility- or study-specific worker files. Dose distributions
are presented in Table 4 for workers monitored and not
monitored for radiation at Rocketdyne/AI by period of
employment. Overall, 932 (16.1%) of the radiation work-
force had prior exposure elsewhere, and 1,224 (21.1%)
had subsequent exposure after leaving Rocketdyne.
Nearly 32% (or 1,833) of the radiation workers had been
employed and monitored for radiation at other facilities.
As seen in Fig. 8, including the radiation dose received
elsewhere had a noticeable influence on the distribution
of doses by increasing the number of workers with
relatively high cumulative doses. Based on the Rocket-
dyne dose only, 231 workers received greater than 50
mSv, but based on all dose information available on
external doses, the number who exceeded 50 mSv over
their career increased to 331, or by 43%.
The amount of misclassification of worker occupa-
tional radiation dose can be seen in Table 5, which presents
a cross-tabulation of external dose received only at Rock-
etdyne/AI by total career dose from all facilities for the
5,743 radiation workers monitored for external radiation at
Rocketdyne/AI. The number of workers in the highest
category (200 mSv) increased from 45 to 69 (or 53%), the
100 –199 mSv category increased from 58 to 100 (or 72%),
whereas the zero dose category decreased from 693 to 601
(or 13%). It can be seen that some workers classified in a
relatively low dose category based on their Rocketdyne/AI
experience received total career doses that placed them in
the highest dose category.
Cumulative organ doses from internal exposures
There were 292 workers whose bioassay measure-
ments or other internal monitoring data indicated that the
committed equivalent dose to at least one tissue might
exceed 10 mSv, which was the criterion for development
of a detailed internal exposure scenario and comprehen-
sive individual organ dose estimates. Organ doses for 16
tissues were estimated for each calendar year after intake
up through the year 1999 (Table 6). The “remainder” in
Table 6 represents the doses received by all other organs
and tissue not explicitly included in the dose computation
assessment, and might be considered a soft tissue dose.
Cumulative organ doses were computed by summing the
annual doses up through 1999 for those alive and to the
date of death for those who died. Consistent with the
major route of exposure being from inhalation, the lung
(mean 106 mSv) and respiratory lymph nodes (mean 300
mSv) had the highest levels. There were 49 workers with
cumulative lung doses exceeding 100 mSv (maximum
3,560 mSv). The major contributor to lung dose was
uranium aluminide. A small number of workers received
cumulative doses exceeding 100 mSv to the bone surface
and to the liver. Plutonium and thorium were the main
contributors to bone surface dose (maximum 5,890
mSv). Two workers received greater than 50 mSv to the
testes and one worker received greater than 10 mSv to the
kidney. Other than for the lung, bone surface, liver, and
perhaps kidney, the doses to the other organs were small
in comparison to the whole-body dose received from
external radiation. Overall, most workers monitored for
internal radiation (1,940 or 86.9%) had negligible in-
takes.
Combining external whole-body doses with internal
organ doses
As is evident from Table 6, the dose distributions
associated with radionuclide intakes for specific organs
are substantially different. These differences are due to
Fig. 7. Urinary data for a worker exposed to uranium before
starting work at Rocketdyne/AI in the early 1960’s.
421Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
the wide range of radionuclides (at least 12) contributing
to internal doses with their different chemistries and
solubility properties. As seen, these different dose distri-
butions are not proportionately related. Accordingly, our
epidemiologic analyses used the dose distribution for
each specific organ, i.e., the external whole-body dose
was added to the internal organ dose for specific cancer
site analyses. The internal dose contribution is substantial
for the lung and may be important for bone surface, liver,
and kidney. For all other organs, few workers received
greater than 5 mSv cumulative internal dose and the
internal dose contribution was small in comparison with
the whole-body external dose received.
Although the number of workers monitored for
intakes of radionuclides was 2,232 or 38% of the radia-
tion workforce, it was only the 292 (5.0%) with relatively
high internal intakes that contributed to the worker doses
as estimated in the present study. For these 292 workers,
dose to the lung was of prime importance, and there were
few workers with relatively high doses to any other
organ. The contribution of internal emitters to lung dose
was meaningful (mean 106 mSv, maximum 3,560 mSv)
even in comparison with the much larger numbers
exposed to external radiation (Fig. 8). Adding the lung
dose from internal emitters to that received from external
exposures increased the number of Rocketdyne workers
who received greater than 50 mSv from 231 (Rocketdyne
only) to 427 (or 84.8%). For all sources of external
radiation, the number who received greater than 50 mSv
increased from 334 to 427 (or 27.8%). Because it was not
possible to compute internal doses for radionuclide
intakes received at facilities other than at Rocketdyne,
other than for a few individuals, it is likely that organ
doses will be underestimated in general within this
population, but especially for the lung.
Occupational exposure received elsewhere
Table 7 presents the average and range of external
and internal radiation doses received at Rocketdyne/AI
and at other places of employment. The total collective
dose from all sources of penetrating radiation was 78.4
person-Sv, of which 20.8 person-Sv (or 26.5%) was
received during employment other than at Rocketdyne.
Fig. 8 provides a visual representation of the influ-
ence on cumulative lung dose from exposures experi-
enced at Rocketdyne, at other facilities, and from internal
radionuclides. Adding the dose received elsewhere and
the internal emitter dose increased the number of high
dose exposures (200 mSv) from 45 to 109, or by a
factor of 2.4 (Table 4). The mean dose to the lung for the
entire population nearly doubled, increasing from 10.0
mSv to 19.0 mSv (Table 7), as did the person-Sieverts
(from 57.4 to 109.4 person-Sv). Just over 5% of workers
(n 292) monitored for internal radiation contributed
disproportionally (28.3%) to the population lung dose.
Neutrons
Nearly 10% (or 584) of the Rocketdyne/AI radiation
workers were monitored for neutron exposures while at
Rocketdyne, and another 1.4% (or 81) were monitored
for neutron exposures elsewhere. However, only about
half (363 or 54%) had positive measurements and there
Table 4. Number of workers monitored and not monitored for radiation at Rocketdyne by cumulative external radiation
dose received before, during and after Rocketdyne employment, neutron dose, internal lung dose, and total career dose.
a
Employment period
Cumulative dose (mSv)
0 5 5- 10- 50- 100- 200 Total
Nonradiation workers 639 718 54 49 11 6 0 1,477
Before and after Rocketdyne
Radiation workers
External photon radiation dose
Before Rocketdyne 105 482 97 165 39 30 14 932
While at Rocketdyne 693 3,231 663 925 128 58 45 5,743
b
After Rocketdyne 744 369 53 41 12 5 0 1,224
TOTAL career, external photons
a
608 3,155 646 1,009 166 97 68 5,749
Total career, neutron dose 303 314 35 11 2 0 0 665
Total career, external (photon &
neutron)
605 3,149 651 1,012 165 100 69 5,751
Internal lung dose 0 24 4 178 37 15 34 292
Total lung dose (all external and
internal)
a
604 3,079 609 1,039 203 113 109 5,756
c
a
Note that the columns do not sum to the “total career” dose. This is because when the radiation dose received elsewhere is added to
the Rocketdyne dose for individual workers, they can move into a higher dose category. For example, a worker with 7 mSv at
Rocketdyne might have received 4 mSv elsewhere and his total career dose of 11 mSv would shift him from the “5-” mSv category
into the “10-” mSv category.
b
58 workers in the radiation cohort who were monitored only for internal radiation are not included in this total.
c
45 workers in the radiation cohort who were monitored only for internal radiation, received no neutron monitoring, and did not have
their internal radiation dose modeled (i.e., did not meet 10 mSv internal dose threshold) are not included in this total.
422 Health Physics May 2006, Volume 90, Number 5
were only 13 workers with cumulative exposures over 10
mSv and none over 100 mSv. In contrast, more than 400
of these same workers had cumulative external exposures
of over 10 mSv. Thus the neutron contribution, if validly
captured, appears negligible in comparison with the
much higher external exposures, contributing less than
3% of the total dose of workers monitored for neutrons
and less than 0.1% of the total dose of all radiation
workers. In general, there is a concern that neutron doses
are underestimated among early radiation workers (Gil-
bert and Fix 1995), either because there was no monitor-
ing for neutrons, the monitoring devices were not accu-
rate, or the mix of thermal and fast neutrons was
unknown and difficult to quantify. It appears that while
Fig. 8. Distribution of workers by lung dose and source of occupational exposure: external (EXT) photon exposure at
Rocketdyne/AI, total career external photon exposure at Rocketdyne/AI and all other facilities, all external (photon and
neutron exposure) at Rocketdyne/AI and elsewhere, all external (photon and neutron exposure) and internal (INT) dose
from Rocketdyne/AI and all other facilities.
Table 5. Cross-tabulation of external dose received only at Rocketdyne by total career dose from all facilities for the
5,743 radiation workers monitored for external radiation at Rocketdyne.
External dose (mSv) received during employment at Rocketdyne only
0 0−5 5- 10- 50- 100- 200 Total
Total career external dose (mSv) 0
601
601
0−5 78
3,067
3,145
5- 6 78
567
651
10- 7 61 76
868
1,012
50- 1 16 7 37
104
165
100- 0 8 9 14 20
49
100
200- 0 1 4 6 4 9
45
69
Total 693 3,231 663 925 128 58 45 5,743
423Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
the number of workers monitored for neutrons is not
inconsequential, perhaps 10% of all workers, the mea-
sured doses per se do not suggest a serious problem for
the Rocketdyne/AI workforce.
DISCUSSION
In general, occupational studies of radiation workers
must be interpreted carefully because (1) occupational
dose received at other nuclear or radiation facilities may
be missed; (2) dose from internally deposited radionu-
clides may be undetermined; (3) dose from neutrons may
be unknown; (4) film badge or thermoluminescent do-
simeter (TLD) exposures are imperfect measures of
organ doses; (5) the dose from natural background
radiation (about 170 mSv in 70 years) may be greater
than the occupational dose; (6) medical x-ray exposures
may be substantial and are rarely available; (7) other
occupational and nonoccupational carcinogens (such as
tobacco use) are usually not considered; (8) selection
biases associated with entry into the work force and
continued employment may be likely; and (9) ascertain-
ment bias is possible if the working population receives
better medical care and more accurate cancer diagnoses
recorded on death certificates than the general population
(Greenwald et al. 1981; Gilbert and Fix 1995; UN-
SCEAR 2000; Daniels et al. 2005; Daniels and
Schubauer-Berigan 2005). The current paper addresses
directly several of these issues within a comprehensive
dosimetry assessment of the Rocketdyne/AI radiation
workforce: Does the dose distribution materially change
when doses received at other facilities are included?
Does the dose distribution materially change when organ
doses from internal radionuclides are added? Does the
dose distribution materially change when doses from
neutrons are taken into account? We were not able to
address directly several issues of potential importance,
such as natural background radiation or exposure to
technically enhanced radioactive materials or medical x
rays, because of the absence of available exposure
documents.
Table 6. Cumulative organ doses for 16 tissues from internal exposures for the 292 Rocketdyne workers with the highest
radionuclide intake. Dose is estimated for 12 radionuclides up through 1999 for those alive and to the date of death for
those who died.
Organ (radionuclide)
a
Cumulative doses (mSv)
Dose characteristics Dose categories
Mean Median Range 1 1- 5- 10- 50- 100- 200- 500
Bladder 0.2 0.1 0−9 284 5 3 0 0 0 0 0
Bone surface (
239
Pu,
232
Th,
90
Sr)
68.7 3.4 0−5,742 31 150 47 38 11 7 4 4
Brain 0.2 0.1 0−9 284 5 3 0 0 0 0 0
Breast 0.2 0.1 0−9 284 5 3 0 0 0 0 0
Colon 0.3 0.1 0−10 283 6 3 0 0 0 0 0
Esophagus 0.2 0.1 0−9 284 5 3 0 0 0 0 0
Kidney (U Type M UA1
x
)
2.6 1.1 0−58 128 136 15 12 1 0 0 0
Liver (
239
Pu)
14.0 0.5 0−1,246 211 51 6 19 1 0 1 3
Lung (UA1
x
U Type M)
106 24.4 0−3,560 7 17 4 178 37 15 19 15
Respiratory lymph nodes 300 4.1 0−16,736 25 133 31 20 15 17 16 35
Red marrow 0.4 0.0 0−18 279 7 3 3 0 0 0 0
Stomach 0.2 0.1 0−9 284 5 3 0 0 0 0 0
Testes (
239
Pu)
1.0 0.1 0−78 275 13 0 2 2 0 0 0
Thyroid 0.2 0.1 0−9 284 5 3 0 0 0 0 0
Remainder 0.2 0.1 0−9 284 5 3 0 0 0 0 0
a
The radionuclide/s typically contributing to the highest organ doses are shown in parentheses.
Table 7. Average and range of external penetrating dose (photons and neutrons) and internal lung dose (mSv) for
workers monitored for radiation at Rocketdyne by period of employment.
Period of employment N
Mean
(mSv)
Maximum
(mSv) Person-Sv
Before Rocketdyne, photons 932 19.0 999 17.7
While at Rocketdyne, photons 5,743 10.0 500 57.4
After Rocketdyne, photons 1,224 2.5 159 3.1
Total career, all periods, photons 5,749 13.5 1,005 77.6
Total career, all periods, neutrons 665 1.2 56 0.8
Internal lung dose 292 106.2 3,560 31.0
Total lung dose, all external and internal 5,756 19.0 3,577 109.4
424 Health Physics May 2006, Volume 90, Number 5
Major conclusions are that doses received at other
facilities are important and meaningfully increase the
number of workers with relatively high career doses; that
a small number of workers exposed to internal radionu-
clides can significantly alter the shape of the dose
distribution for certain organs; and that neutron expo-
sures apparently were low and had little influence on the
shape of the dose distributions. Other findings of note are
that nearly 4% of non-radiation workers were monitored
for radiation elsewhere during their careers; existing
biokinetic models for radionuclides are not always suf-
ficient to compute organ doses for uncommon radioac-
tive compounds; and that exhaustive attempts to collect
lifetime career doses for radiation workers may still
underestimate exposures because of the difficulty in
capturing all occupational doses, e.g., doses for workers
employed by the U.S. Navy were not available except for
a few workers, and because of the difficulty in determin-
ing, much less computing, organ doses, for radionuclide
intakes and neutron exposures received elsewhere.
As mentioned above, this paper does not address all
the possible uncertainties in dose reconstruction but
focuses on the three that potentially could result in
spurious epidemiologic conclusions, i.e., occupational
doses received elsewhere, inadequate handling of organ
doses from internal radionuclide intakes, and ignoring
neutron exposures. Other assumptions commonly made
in epidemiologic investigations of radiation workers deal
with low dose recording conventions, missing data,
exposure geometry, dosimetry system, and type of re-
corded data. There were few external dosimeter data
values that were recorded as “less than some small
number,” and these and “zeros” were treated as if no dose
were received during the reporting period, usually every
3 mo. Because we had seven overlapping sources of
external radiation, the likelihood of missing a significant
amount of dosimetry reports appears small. For external
doses, no geometric adjustments were made (i.e.,
whether the exposure was AP, PA, lateral or isotropic),
and we assumed that the measured dose was a good
approximation of the organ dose as is done in practically
all occupational studies of radiation workers. The meth-
ods of measuring external dose changed over the years,
from film badge to TLD to OSL (optically stimulated
luminescence), and exposures were reported in a variety
of endpoints (such as shallow dose, penetrating dose,
beta dose, photon dose, neutron dose). For the doses used
in the epidemiologic investigation, we did not assume
that the measured dose differed by type of dosimetry
system used, and in all cases penetrating dose was used
as the best estimate for organ dose. The other reported
doses were not used, except the neutron dose when
recorded.
The radiation dosimetry approach taken in the cur-
rent study differs from the one used in the previous study,
making direct comparisons problematic (Ritz et al. 1999,
2000; Morgenstern and Ritz 2001). Differences include
the selection criteria for cohort members; the efforts
made to capture and incorporate radiation doses received
elsewhere, the inclusion of neutron doses, and the esti-
mation of dose following the intake of radionuclides. Our
study population is larger by about 1,200 radiation
workers than the previous study for reasons that are not
entirely clear since essentially the same Rocketdyne files
formed the basis of the study cohorts. There were
differences in selection criteria in that we included
workers employed between 1948 –1999 but excluded
those who worked less than 6 mo, whereas the previous
study included all workers monitored for radiation be-
tween 1950 –1993, and made no exclusions based on
duration of employment. Most of the differences, how-
ever, appeared in our including more workers with low
doses of 5 mSv. We also made an intensive effort to
obtain additional radiation doses received at installations
other than at Rocketdyne and incorporated these doses in
the epidemiologic analyses. The previous study recorded,
but did not use, the prior employment doses for over 400
workers from available notices in the Rocketdyne radia-
tion folders, whereas we found 932 workers with prior
doses from the record linkage approaches described as
well as an additional 1,224 instances of workers who
were monitored for radiation after leaving Rocketdyne/
Atomics International. Nearly 10% of the radiation
cohort was monitored for neutron exposures at Rocket-
dyne and another 2% were monitored for neutron expo-
sures elsewhere. We incorporated these neutron doses in
each worker’s total career dose, whereas they were
excluded in the previous study. The last important
difference is in the approach to internal dosimetry. We
used the latest ICRP biokinetic models for organ dose
determinations for 12 radionuclides, including three
models provisionally adopted for upcoming ICRP reports
and one model that we independently developed for
uranium aluminide (Leggett et al. 2005). We computed
doses for 16 different organs or tissues up until the time
of death or to the end of study in 1999. The previous
study considered only about 5 radionuclides, used out-
dated ICRP biokinetic models, and only calculated cu-
mulative equivalent doses to the lung and no other organ.
The estimated lung doses were used in the previous study
to approximate the doses to all other organs and tissues,
but this does not appear from our more detailed estimates
to be a valid assumption (Table 6).
The computation of organ doses following intake of
radionuclides for use in an epidemiologic study differs
from the normal procedures used in radiation protection.
425Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
An epidemiologic investigation requires a radiation dose
to a specific organ that is received up to some period of
time prior to the diagnosis of a malignancy or an
equivalent time for those without disease. For radiation
protection, the committed effective dose is computed
following an intake that projects the effective dose over
a period of 50 y based on administratively defined tissue
and radiation weighting factors (ICRP 1991). In this
study, a major effort was made to compute annual organ
doses from inhaled or ingested radionuclides for 16
specific organs or tissues, and the ICRP models were
modified accordingly. Effective dose is not appropriate
for epidemiologic analyses and was not computed (Cox
and Kellerer 2003).
The computation of organ doses following intake of
radionuclides for use in an epidemiologic study is in
principle the same as the computation performed for an
individual for compensable considerations, although the
completeness of the bioassay and dosimetry data available
for Rocketdyne/AI workers is rather exceptional and there
was little uncertainty in the fact of exposure or the relative
intake. Nonetheless, we did not perform the laborious dose
reconstruction computations when the organ doses were
clearly inconsequential, i.e., for workers with committed
equivalent doses indicated by screening criteria to be less
than 10 mSv to the organ with the highest internal dose.
Similar considerations might be made for radiation workers
or atomic veterans when the available dosimetry is sparse or
missing (Office of Workers’ Compensation 2005; NRC
2003, 2005). For individuals with missing dosimetry but
whose exposure group is reasonably well known, the
maximum known intake for the individuals in the exposure
group could be assigned to those individuals with uncertain
exposure. If based on this “maximum assumption” the
computed organ doses are substantially less than what
would be required for compensation, then a dose recon-
struction would not be necessary. If the maximum dose
results in a probability of causation (assigned share) value
that was above a compensable level, then the individuals in
the group could receive compensation based on the legal
mandate to err on the side of compassion and for the benefit
of the exposed individual.
Those employed in the early years of the atomic age
often worked at several facilities during their career
making it difficult to collect dose information received
elsewhere, especially from subsequent places of employ-
ment, which are rarely recorded in worker files. Expo-
sures to inhaled or ingested radionuclides are difficult to
quantify in terms of organ doses even when complete
bioassay measurements and sophisticated biokinetic
models are available. Neutron exposures were often not
recorded or were poorly estimated. Workers prior to
about 1970 could have received occupational x-ray
examinations, including photofluorography, which could
be large and might even be considered part of their
occupational exposure (Daniels et al. 2005). The conse-
quences of these limitations and uncertainties could lead
to a serious underestimation of dose for workers and an
overestimation of the risk at low doses, or perhaps to
such serious misclassification that results might not be
interpretable. Until recently, few ways were available to
address these problems, although sensitivity analyses
have been conducted to estimate the magnitude of
possible biases (Gilbert and Fix 1995).
Over the years, a wide range of approaches have
been taken by various investigators to deal with transfer
doses, i.e., doses received at facilities other than the one
being studied, neutron exposures, and doses from inter-
nally deposited radionuclides. Transfer doses received
prior to employment at the study facility have been
included in some studies when known (Gilbert et al.
1993a and b; McGeoghegan and Binks 2000a and b,
2001; McGeoghegan et al. 2003), excluded in some
studies when known (Ritz et al. 1999; Wing et al. 2004),
but are usually not considered except perhaps in nation-
wide dose registry studies where it is assumed that most,
but not all, radiation installations have contributed to the
database (Muirhead et al. 1999; Sont et al. 2001; Iwasaki
et al. 2003). Few investigations have been able to
determine radiation exposures for workers who have left
a particular facility. More than half (500/932) of the
Rocketdyne/AI workers found to have been monitored
for radiation prior to joining Rocketdyne had no mention
of this previous radiation experience in their folders, and,
not surprisingly, practically all of the subsequent expo-
sures received elsewhere by over 21% (1,224/5,801) of
the workforce were not mentioned.
Except for large-scale epidemiologic studies of plu-
tonium workers in the United Kingdom and in Russia,
few if any epidemiologic studies have attempted to
compute organ doses following the ingestion or inhala-
tion of radionuclides. Although internal radiation was not
considered in the early analyses of Sellafield workers
(Douglas et al. 1994), recent comprehensive reports have
incorporated organ-specific doses from plutonium (Rid-
dell et al. 2000; Riddell 2002), and some analyses were
based on the total dose from external exposure and
internal emitters (Omar et al. 1999). Individual assess-
ments had been made for perhaps 20% of the workforce
because of special circumstances such as statutory re-
quirements, compensation, and operational protection.
For the remaining workers, an automated assessment
program was used to characterize doses based on avail-
able urinalysis results and other factors. Some studies
have considered the fact that workers were monitored (or
had the potential to be monitored) for radionuclides in
426 Health Physics May 2006, Volume 90, Number 5
the analyses (Frome et al. 1997; Beral et al. 1988; Fraser
et al. 1993; Carpenter et al. 1998), others were aware of
radionuclide exposures but apparently did not sort them
out of the analyses (Wiggs et al. 1991; Inskip et al. 1987;
Carpenter et al. 1994; Rooney et al. 1993; Cardis et al.
1995; Muirhead et al. 1999; Dupree-Ellis et al. 2000;
McGeoghegan and Binks 2000a and b), and others
excluded workers monitored for internal radionuclides
(Gilbert et al. 1993a, and b; Cardis et al. 2005).
Comprehensive dosimetry continues for Mayak
workers in Russia exposed to very high levels of pluto-
nium, i.e., body burdens of the order of 4 6 kBq, and
organ doses have been computed for lung, liver and bone
surfaces and other organs. Both body burdens and organ
doses from plutonium and external doses have been
analyzed in a variety of ways (Koshurnikova et al. 1998,
2000; Gilbert et al. 2000; IARC 2001; Kreisheimer et al.
2003; Shilnikova et al. 2003). In the United States,
dose-response analyses have been conducted of internal
intakes of plutonium, but body burdens and not organ
doses were used (Wiggs et al. 1994; Wilkinson et al.
1987; Voelz et al. 1997). In contrast to the Mayak
studies, the body burdens were quite low in the U.S.
worker studies, most on the order of 100 Bq, or about 50
times lower than those in Russian workers. The previous
study of Rocketdyne workers computed lung dose equiv-
alents for five radionuclides based on bioassay measure-
ments (Ritz et al. 1999). One study of Hanford plutonium
workers based analyses on job titles alone and not on
available bioassay data (Wing et al. 2004). A recent study
of Rocky Flats plutonium workers estimated annual
equivalent lung doses based on urinary bioassay data and
lung counts using the computer code CINDY (Brown et
al. 2004).
In our conduct of the dose compilations, we were
able to address several important issues regarding the
estimation of internal exposure for epidemiological set-
tings. In an epidemiological scheme relying on organ
dose as an analogue for risk, we demonstrated that body
burdens do not provide a consistent indicator of such
risk. Since the chemical properties of the radioactive
material determines its retention in the body and addi-
tionally influences the nuclide’s behavior in specific
organs of concentration, and since organ dose is directly
dependent upon these physiological processes (e.g.,
Priest 1989), the only valid way to estimate dose to an
organ was through modeling. We found that for the
Rocketdyne workers, the computed body burden would
result in widely divergent doses between organs, i.e., the
lung generally received doses significantly higher than
other evaluated organ sites and, coupled with the differ-
ent types of radionuclides contributing to organ doses,
there was no proportional relationship between dose to
lung and dose to other organs as assumed in the previous
study (Ritz et al. 2000). Thus, an implicit assumption that
a body burden imparted a universal or linear risk to all
organs of the body was demonstrated to be unsupport-
able. We also found that most persons monitored for
radionuclides did not have positive readings, thus making
categorization by job title or other work activity-based
schemes unreliable.
Neutron exposures have been inconsistently handled
also. Some studies recorded and analyzed neutron doses
(Inskip et al. 1987), others excluded neutron-exposed
workers in some analyses (Cardis et al. 1995, 2005), and
others ignored the neutron dose (Ritz et al. 1999). Other
studies noted neutron exposures but did not address them
in the analyses (Gilbert et al. 2000, 1993b). Although
nearly 12% of the Rocketdyne workforce was monitored
for neutron exposures, the contribution to total dose was
not great. It is generally agreed, however, that neutron
exposures were poorly estimated in the early years of the
atomic age (Gilbert 1993b).
The problem associated with excluding workers for
whatever reason can be a reduction in study power,
which might be further exacerbated if excluded workers
are those likely to have received relatively high expo-
sures. Validity could be affected, however, if workers
with radionuclide intakes, neutron exposures, and doses
received at other facilities are inappropriately handled.
We attempted to address these uncertainties by seeking
exposure histories from all sources available, computing
organ-specific annual doses following internal radionu-
clide intakes, and recording and incorporating all avail-
able neutron exposure data.
In conclusion, while the Rocketdyne/AI radiation
study is limited because of the relatively small number
of workers and narrow range of organ doses, it
nonetheless illustrates some of the pitfalls in conduct-
ing occupational studies of low doses where seemingly
minor uncertainties can have a substantial impact on
the estimation of population doses. Exposures re-
ceived elsewhere increased the collective dose by 38%
and the contribution of radionuclide intakes to certain
organs greatly increased the number of workers with
high doses; e.g., the lung dose person-Sv was in-
creased by 55%. While the number of workers moni-
tored for neutron exposures was not small, the actual
neutron dose recorded did not appreciably affect the
dose distributions. A small percentage (3.6%) of
non-radiation workers were also found to have been
monitored for radiation elsewhere. These sources of
bias should be considered when interpreting results
from radiation studies of workers.
427Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
Acknowledgments—We thank the U.S. Department of Energy (Nimi Rao),
the Nuclear Regulatory Commission (Rosemary Hogan and Sheryl Bur-
rows), SAIC (Derek A. Hagemeyer), the U.S. Army Radiation Standards
and Dosimetry Laboratory (William S. Harris, Jr., CHP) and the U.S. Air
Force Radiation Surveillance Division, Air Force Institute for Operational
Health (Gerald Achenbach and Mike Klueber) for providing linkages with
their respective dosimetry files. We are grateful for the helpful advice
provided by Barbara Brooks (CEDR Program Coordinator, DOE), James
G. Barnes, CHP (Radiation Safety Officer, Rocketdyne, The Boeing
Company), and Judy McLaughlin (Rocketdyne, The Boeing Company).
The study was supported in part by a competitive contract from The Boeing
Company and was conducted with cooperation from the United Automo-
bile, Aerospace and Agricultural Implement Workers of America (UAW)
and from Oak Ridge Institute for Science and Education (ORISE) through
an interagency agreement with the U.S. DOE. ORISE is managed by Oak
Ridge Associated Universities under DOE contract number DE-AC05-
00OR22750. The results presented herein represent the conclusions and
opinions solely of the authors. Its publication does not imply endorsement
by The Boeing Company, the UAW, or any of the acknowledged agencies.
REFERENCES
Beral V, Fraser P, Carpenter L, Booth M, Brown A, Rose G.
Mortality of employees of the Atomic Weapons Establish-
ment, 1951– 82. Br Med J 297:757–770; 1988.
Bland RD. A parametric study of ion-plated aluminum coatings
on uranium. Electrochem Tech 6:272–278; 1968.
Boice JD Jr. Leukemia risk in thorotrast patients. Radiat Res
134:224 –233; 1993.
Boice JD Jr. Ionizing radiation. In: Schottenfeld D, Fraumeni
JF, Jr, eds. Cancer epidemiology and prevention. New
York: Oxford University Press; 2006: 256 –293.
Brown SC, Schonbeck MF, McClure D, Baron AE, Navidi
WC, Byers T, Ruttenber AJ. Lung cancer and internal lung
doses among plutonium workers at the Rocky Flats Plant: a
case-control study. Am J Epidemiol 160:163–172; 2004.
Buddery JH, Clark ME, Pearce RJ, Stobbs JJ. The development
and properties of an oxidation-resistant coating for uranium.
J Nucl Mat 13:169 –181; 1964.
Cardis E, Gilbert ES, Carpenter L, Howe G, Kato I, Armstrong
BK, Beral V, Cowper G, Douglas A, Fix J, Kaldor J, Lave
C, Salmon L, Smith PG, Voelz GL, Wiggs LD. Effects of
low doses and low dose rates of external ionizing radiation:
cancer mortality among nuclear industry workers in three
countries. Radiat Res 142:117–132; 1995.
Cardis E, Vrijheid M, Blettner M, Gilbert E, Hakama M, Hill
C, Howe G, Kaldor J, Muirhead CR, Schubauer-Berigan M,
Yoshimura T, Bermann F, Cowper G, Fix J, Hacker C,
Heinmiller B, Marshall M, Thierry-Chef I, Utterback D,
Ahn Y-O, Amoros E, Ashmore P, Auvinen A, Bae J-M,
Bernar Solano J, Biau A, Combalot E, Deboodt P, Diez
Sacristan A, Eklof M, Engels H, Engholm G, Gulis G,
Habib R, Holan K, Hyvonen H, Kerekes A, Kurtinaitis J,
Malker H, Martuzzi M, Mastauskas A, Monnet A, Moser M,
Pearce MS, Richardson DB, Rodriguez-Artalejo F, Rogel
A, Tardy H, Telle-Lamberton M, Turai I, Usel M, Veress K.
Risk of cancer after low doses of ionising radiation: Retro-
spective cohort study in 15 countries. BMJ 331:77; 2005.
Carpenter L, Higgins CD, Douglas A, Fraser P, Beral V, Smith
P. Combined analysis of mortality in three United Kingdom
nuclear industry workforces, 1946 –1988. Radiat Res
138:224 –238; 1994.
Carpenter LM, Higgins, Douglas AJ, Maconochie NES, Omar
RZ, Fraser P, Beral V, Smith PG. Cancer mortality in
relation to monitoring for radionuclide exposure in three
UK nuclear industry workforces. Br J Cancer 78:1224
1232; 1998.
Cox R, Kellerer AM. A current view in radiation weighting
factors and effective dose. Oxford: Pergamon Press; ICRP
Pub 92; Ann ICRP 33:1– 4; 2003.
Daniels RD, Schubauer-Berigan MK. Bias and uncertainty of
penetrating photon dose measured by film dosemeters in an
epidemiological study of US nuclear workers. Radiat Pro-
tect Dosim 113:275–289; 2005.
Daniels RD, Kubale TL, Spitz HB. Radiation exposure from
work-related medical x-rays at the Portsmouth Naval Ship-
yard. Am J Ind Med 47:206 –216; 2005.
Douglas AJ, Omar RZ, Smith PG. Cancer mortality and
morbidity among workers at the Sellafield plant of British
Nuclear Fuels. Br J Cancer 70:1232–1243; 1994.
Dupree-Ellis E, Watkins J, Ingle JN, Phillips J. External
radiation exposure and mortality in a cohort of uranium
processing workers. Am J Epidemiol 152:91–95; 2000.
Fraser P, Carpenter L, Maconochie N, Higgins C, Booth M,
Beral V. Cancer mortality and morbidity in employees of
the United Kingdom Atomic Energy Authority, 1946 86.
Br J Cancer 67:615– 624; 1993.
Frome EL, Cragle DL, Watkins JP, Wing S, Shy CM, Tank-
ersley WG, West CM. A mortality study of employees of
the nuclear industry in Oak Ridge, Tennessee. Radiat Res
148:64 80; 1997.
Fry SA, Dupree EA, Sipe AH, Seiler DL, Wallace PW. A study
of mortality and morbidity among persons occupationally
exposed to 50 mSv in a year: phase I, mortality through
1984. Appl Occup Environ Hyg 11:334 –343; 1996.
Gilbert ES. Accounting for errors in dose estimates used in
studies of workers exposed to external radiation. Health
Phys 74:22–29; 1998.
Gilbert ES. Invited commentary: studies of workers exposed to
low doses of radiation. Am J Epidemiol 153:319–322;
discussion 323–324; 2001.
Gilbert ES, Fix JJ. Accounting for bias in dose estimates in
analyses of data from nuclear worker mortality studies.
Health Phys 68:650 660; 1995.
Gilbert ES, Omohundro E, Buchanan JA, Holter NA. Mortality
of workers at the Hanford site: 1945–1986. Health Phys
64:577–590; 1993a.
Gilbert ES, Cragle DL, Wiggs LD. Updated analyses of
combined mortality data for workers at the Hanford Site,
Oak Ridge National Laboratory, and Rocky Flats Weapons
Plant. Radiat Res 136:408 421; 1993b.
Gilbert ES, Fix JJ, Baumgartner WV. An approach to evalu-
ating bias and uncertainty in estimates of external dose
obtained from personal dosimeters. Health Phys 70:336
345; 1996.
Gilbert ES, Koshurnikova NA, Sokolnikov M, Khokhryakov
VF, Miller S, Preston DL, Romanov SA, Shilnikova NS,
Suslova KG, Vostrotin VV. Liver cancers in Mayak work-
ers. Radiat Res 154:246 –252; 2000.
Giles RD, Tavender LE. Preparation and properties of UAlx
coatings formed on uranium via the electrophoretic deposi-
tion of aluminium powder. J Nucl Mat 24:129 –140; 1967.
Greenwald P, Friedlander BR, Lawrence CE, Hearne T, Earle
K. Diagnostic sensitivity bias—an epidemiologic explana-
tion for an apparent brain tumor excess. J Occup Med
23:690 694; 1981.
Inskip H, Beral V, Fraser P, Booth M, Coleman D, Brown A.
Further assessment of the effects of occupational radiation
exposure in the United Kingdom Atomic Energy Authority
mortality study. Br J Ind Med 44:149–160; 1987.
International Agency for Research on Cancer. IARC mono-
graphs on the evaluation of carcinogenic risks to humans.
428 Health Physics May 2006, Volume 90, Number 5
Vol 78. Ionizing radiation, part 2: Some internally deposited
radionuclides. Lyon, France: IARC; 2001.
International Commission on Radiological Protection. Limits
for intakes of radionuclides by workers. Part 1. Oxford:
Pergamon Press; ICRP Publication 30; Ann. ICRP, Vol. 2,
No. 3/4; 1979.
International Commission on Radiological Protection. 1990
recommendations of the International Commission on Ra-
diological Protection. Oxford: Pergamon Press; Publication
60; Ann. ICRP, Vol. 21; 1991.
International Commission on Radiological Protection. Age
dependent doses to members of the public from intake of
radionuclides. Part 2. Oxford: Pergamon Press; ICRP Pub-
lication 67; 1993.
International Commission on Radiological Protection. Dose
coefficients for intakes of radionuclides by workers: re-
placement of ICRP Publication 61. Oxford: Pergamon
Press; ICRP Publication 68; Ann. ICRP, Vol. 24; 1994a.
International Commission on Radiological Protection. Human
respiratory tract model for radiological protection. Oxford:
Pergamon Press; ICRP Publication 66; Ann. ICRP, Vol. 26,
No. 1; 1994b.
International Commission on Radiological Protection. Age-
dependent doses to members of the public from intake of
radionuclides. Part 3. Ingestion dose coefficients. Oxford:
Pergamon Press; ICRP Publication 69; Ann. ICRP 25, No 1;
1995a.
International Commission on Radiological Protection. Age-
dependent doses to members of the public from intake of
radionuclides. Part 4. Inhalation dose coefficients. Oxford:
Pergamon Press; ICRP Publication 71; Ann. ICRP 25, No.
3– 4; 1995b.
International Commission on Radiological Protection. Individ-
ual monitoring for internal exposure of workers. Oxford:
Pergamon Press; ICRP Publication 78; 1997.
Iwasaki T, Murata M, Ohshima S, Miyake T, Kudo S, Inoue Y,
Narita M, Yoshimura T, Akiba S, Tango T, Yoshimoto Y,
Shimizu Y, Sobue T, Kusumi S, Yamagishi C, Matsudaira
H. Second analysis of mortality of nuclear industry workers
in Japan, 1986 –1997. Radiat Res 159:228 –238; 2003.
Koshurnikova NA, Bolotnikova MG, Ilyin LA, Keirim-Markus
IB, Menshikh ZS, Okatenko PV, Romanov SA, Tsvetkov
VI, Shilnikova NS. Lung cancer risk due to exposure to
incorporated plutonium. Radiat Res 149:366 –371; 1998.
Koshurnikova NA, Gilbert ES, Sokolnikov M, Khokhryakov
VF, Miller S, Preston DL, Romanov SA, Shilnikova NS,
Suslova KG, Vostrotin VV. Bone cancers in Mayak work-
ers. Radiat Res 154:237–245; 2000.
Kreisheimer M, Sokolnikov ME, Koshurnikova NA,
Khokhryakov VF, Romanow SA, Shilnikova NS, Okatenko
PV, Nekolla EA, Kellerer AM. Lung cancer mortality
among nuclear workers of the Mayak facilities in the former
Soviet Union. An updated analysis considering smoking as
the main confounding factor. Radiat Environ Biophys
42:129 –135; 2003.
Le Claire AD, Bear IJ. The interdiffusion of uranium and
aluminum. J Nucl Energy 2:229 –242; 1956.
Leggett RW, Eckerman KF. A systemic biokinetic model for
polonium. Sci Total Environ 275:109 –125; 2001.
Leggett RW, Eckerman KF, Boice JD, Jr. A respiratory model
for uranium aluminide based on occupational data. J Radiol
Prot 25:405– 416; 2005.
McGeoghegan D, Binks K. The mortality and cancer morbidity
experience of workers at the Capenhurst uranium enrich-
ment facility 1946 –95. J Radiol Prot 20:381– 401; 2000a.
McGeoghegan D, Binks K. The mortality and cancer morbidity
experience of workers at the Springfields uranium produc-
tion facility, 1946 –95. J Radiol Protect 20:111–137; 2000b.
McGeoghegan D, Binks K. The mortality and cancer morbidity
experience of employees at the Chapelcross plant of British
Nuclear Fuels plc, 1955–95. J Radiol Prot 21:221–250;
2001.
McGeoghegan D, Gillies M, Riddell AE, Binks K. Mortality
and cancer morbidity experience of female workers at the
British Nuclear Fuels Sellafield plant, 1946 –1998. Am J Ind
Med 44:653– 663; 2003.
Morgenstern H, Ritz B. Effects of radiaiton and chemical
exposures on cancer mortality among Rocketdyne workers:
a review of three cohort studies. Occup Med 16:219 –237;
2001.
Muirhead CR, Boice JD, Jr, Raddatz CT, Yoder RC. Compar-
ison of dose histories for U.S. nuclear power plant workers,
based on records held by a major dosimetry service com-
pany and on the NRC REIRS database. Health Phys
70:645– 650; 1996.
Muirhead CR, Goodill AA, Haylock RG, Vokes J, Little MP,
Jackson DA, O’Hagan JA, Thomas JM, Kendall GM, Silk
TJ, Bingham D, Berridge GL. Occupational radiation expo-
sure and mortality: second analysis of the National Registry
for Radiation Workers. J Radiol Prot 19:3–26; 1999.
National Council on Radiation Protection and Measurements.
The relative biological effectiveness of radiations of differ-
ent quality. Bethesda, MD: National Council on Radiation
Protection and Measurements; NCRP Report No. 104;
1990.
National Research Council. Committee to Review the Dose
Reconstruction Program of the Defense Threat Reduction
Agency. A review of the dose reconstruction program of the
Defense Threat Reduction Agency. Washington, DC: Na-
tional Academy Press; 2003.
National Research Council. Assessment of the scientific infor-
mation for the Radiation Exposure Screening and Education
Program. Washington DC: National Academies Press;
2005.
Nuclear Regulatory Commission and the Commission of Eu-
ropean Communities. Probabilistic accident consequence
uncertainty analysis. Late health effects uncertainty assess-
ment. Washington, DC: U.S. Nuclear Regulatory Commis-
sion; Luxembourg: Office for Publications of the European
Communities; NUREG/CR-6555; EUR 16774; SAND97-
2322; 1997.
Office of Workers’ Compensation Programs, Employment
Standards Administration, Labor. Performance of functions;
claims for compensation under the Energy Employees
Occupational Illness Compensation Program Act. Interim
final rule; request for comments. Fed Regist 70:33589
33639; 2005.
Omar RZ, Barber JA, Smith PG. Cancer mortality and mor-
bidity among plutonium workers at the Sellafield plant of
British Nuclear Fuels. Br J Cancer 79:1288–1301; 1999.
Priest ND. Alpha-emitters in the skeleton: an evaluation of the
risk of leukaemia following intakes of plutonium 239. In:
Taylor DM, Mays CW, Gerber GB, Thomas RG, eds. Risks
from radium and thorotrast. Report 21. London: British
Institute of Radiology; 1989: 159 –165.
Riddell AE. Advances in the assessment of internal dose for
workforce epidemiological studies. In: Proceedings of
Fourth International Conference on Health Effects of Low
Radiation. Paper 9. Oxford: British Nuclear Energy Society;
2002.
429Dose reconstruction for former Rocketdyne workers
J. D. BOICE ET AL.
Riddell AE, Battersby WP, Peace MS, Strong R. The assess-
ment of organ doses from plutonium for an epidemiological
study of the Sellafield workforce. J Radiol Prot 20:275–286;
2000.
Ritz B, Morgenstern H, Froines J, Young BB. Effects of
exposure to external ionizing radiation on cancer mortality
in nuclear workers monitored for radiation at Rocketdyne/
Atomics International. Am J Ind Med 35:21–31; 1999.
Ritz B, Morgenstern H, Crawford-Brown D, Young B. The
effects of internal radiation exposure on cancer mortality in
nuclear workers at Rocketdyne/Atomics International. En-
viron Health Perspect 108:743–751; 2000.
Rooney C, Beral V, Maconochie N, Fraser P, Davies G.
Case-control study of prostatic cancer in employees of the
United Kingdom Atomic Energy Authority. BMJ
307:1391–1397; 1993.
Shilnikova NS, Preston DL, Ron E, Gilbert ES, Vassilenko EK,
Romanov SA, Kuznetsova IS, Sokolnikov ME, Okatenko
PV, Kreslov VV, Koshurnikova NA. Cancer mortality risk
among workers at the Mayak nuclear complex. Radiat Res
159:787–798; 2003.
Sont WN, Zielinski JM, Ashmore JP, Jiang H, Krewski D, Fair
ME, Band PR, Letourneau EG. First analysis of cancer
incidence and occupational radiation exposure based on the
National Dose Registry of Canada. Am J Epidemiol
153:309 –318; 2001.
Subramanyam D, Notis MR, Goldstein JI. Microstructural
investigation of intermediate phase formation on uranium-
aluminum diffusion couples. Metallurgical Transactions
16A:589 –595; 1985.
Taylor DM, Leggett RW. A generic biokinetic model for
predicting the behaviour of the lanthanide elements in the
human body. Radiat Prot Dosim 105:193–198; 2003.
United Nations Scientific Committee on the Effects of Atomic
Radiation. UNSCEAR 2000 Report to the General Assem-
bly, with Scientific Annexes. Sources and effects of ioniz-
ing radiation. New York: United Nations; E.00.IX.4; 2000.
U.S. Environmental Protection Agency. Cancer risk coeffi-
cients for environmental exposure to radionuclides. Wash-
ington, DC: Environmental Protection Agency; Federal
Guidance Report No. 13, EPA402R99001; 1999.
Voelz GL, Lawrence JNP, Johnson ER. Fifty years of pluto-
nium exposure to the Manhattan Project plutonium workers:
an update. Health Phys 73:611– 619; 1997.
Wiggs LD, Cox-Devore CA, Voelz GL. Mortality among a
cohort of workers monitored for Po-210 exposure: 1944
1972. Health Phys 61:71–76; 1991.
Wiggs LD, Johnson ER, Cox-DeVore CA, Voelz GL. Mortal-
ity through 1990 among white male workers at the Los
Alamos National Laboratory: considering exposures to
plutonium and external ionizing radiation. Health Phys
67:577–588; 1994.
Wilkinson GS, Tietjen GL, Wiggs LD, Galke WA, Acquavella
JF, Reyes M, Voelz GL, Waxweiler RJ. Mortality among
plutonium and other radiation workers at a plutonium
weapons facility. Am J Epidemiol 125:231–250; 1987.
Wing S, Richardson D, Wolf S, Mihlan G. Plutonium-related
work and cause-specific mortality at the United States
Department of Energy Hanford Site. Am J Ind Med 45:153–
164; 2004.
ff
430 Health Physics May 2006, Volume 90, Number 5
    • "Accordingly, assessment of exposure levels was based upon the similarity of exposures in the industrial process combined with expert knowledge and judgement (e.g. from occupational health and safety personnel, as well as the workers under study). Given that comparisons of groups of workers involved in the same work often show similar urinary excretion of radionuclides over an extended period [43], the validity of a JEM approach based upon this assumption has some support. In other industries, some validation has been attempted by, for example, setting aside part of the data prior to JEM development and then using the JEM to estimate the withheld data [91]. "
    [Show abstract] [Hide abstract] ABSTRACT: Any potential health effects of radiation emitted from radionuclides deposited in the bodies of workers exposed to radioactive materials can be directly investigated through epidemiological studies. However, estimates of radionuclide exposure and consequent tissue-specific doses, particularly for early workers for whom monitoring was relatively crude but exposures tended to be highest, can be uncertain, limiting the accuracy of risk estimates. We review the use of job-exposure matrices (JEMs) in peer-reviewed epidemiological and exposure assessment studies of nuclear industry workers exposed to radioactive materials as a method for addressing gaps in exposure data, and discuss methodology and comparability between studies. We identified nine studies of nuclear worker cohorts in France, Russia, the USA and the UK that had incorporated JEMs in their exposure assessments. All these JEMs were study or cohort-specific, and although broadly comparable methodologies were used in their construction, this is insufficient to enable the transfer of any one JEM to another study. Moreover there was often inadequate detail on whether, or how, JEMs were validated. JEMs have become more detailed and more quantitative, and this trend may eventually enable better comparison across, and the pooling of, studies. We conclude that JEMs have been shown to be a valuable exposure assessment methodology for imputation of missing exposure data for nuclear worker cohorts with data not missing at random. The next step forward for direct comparison or pooled analysis of complete cohorts would be the use of transparent and transferable methods.
    Full-text · Article · Mar 2016
    • "@BULLET At the planning stage of the study, all relevant information has to be collected and scenarios of exposure to radiation have to be established on the basis of that information, which includes all available dose records and monitoring data, reviews of the documents from the site files, and interviews of long-time site workers. For the workers that spent their career in several radiation facilities, the relevant records from those facilities have to be recovered in order to estimate the entire radiation exposure history for those workers (e.g., Boice et al. 2006 ). It is useful to prepare a flowchart identifying all steps of the dose assessment for all categories of study subjects. "
    [Show abstract] [Hide abstract] ABSTRACT: The primary aim of the epidemiologic study of one million U.S. radiation workers and veterans [the Million Worker Study (MWS)] is to provide scientifically valid information on the level of radiation risk when exposures are received gradually over time and not within seconds, as was the case for Japanese atomic bomb survivors. The primary outcome of the epidemiologic study is cancer mortality, but other causes of death such as cardiovascular disease and cerebrovascular disease will be evaluated. The success of the study is tied to the validity of the dose reconstruction approaches to provide realistic estimates of organ-specific radiation absorbed doses that are as accurate and precise as possible and to properly evaluate their accompanying uncertainties. The dosimetry aspects for the MWS are challenging in that they address diverse exposure scenarios for diverse occupational groups being studied over a period of up to 70 y. The dosimetric issues differ among the varied exposed populations that are considered: atomic veterans, U.S. Department of Energy workers exposed to both penetrating radiation and intakes of radionuclides, nuclear power plant workers, medical radiation workers, and industrial radiographers. While a major source of radiation exposure to the study population comes from external gamma- or x-ray sources, for some of the study groups, there is a meaningful component of radionuclide intakes that requires internal radiation dosimetry assessments. Scientific Committee 6-9 has been established by the National Council on Radiation Protection and Measurements (NCRP) to produce a report on the comprehensive organ dose assessment (including uncertainty analysis) for the MWS. The NCRP dosimetry report will cover the specifics of practical dose reconstruction for the ongoing epidemiologic studies with uncertainty analysis discussions and will be a specific application of the guidance provided in NCRP Report Nos. 158, 163, 164, and 171. The main role of the Committee is to provide guidelines to the various groups of dosimetrists involved in the MWS to ensure that certain dosimetry criteria are considered: calculation of annual absorbed doses in the organs of interest, separation of low and high linear-energy transfer components, evaluation of uncertainties, and quality assurance and quality control. It is recognized that the MWS and its approaches to dosimetry are a work in progress and that there will be flexibility and changes in direction as new information is obtained with regard to both dosimetry and the epidemiologic features of the study components. This paper focuses on the description of the various components of the MWS, the available dosimetry results, and the challenges that have been encountered. It is expected that the Committee will complete its report in 2016.
    Full-text · Article · Feb 2015
    • "e internal dose to the lung (Checkoway et al. 1988; Dupree et al. 1995; Ritz 1999b; Ritz et al. 2000; Richardson and Wing 2006), lung cancer risk was raised for dose categories from 10 –30 mSv and more. Since few deaths occurred in the high dose categories, the risk increase did not reach statistical significance and a clear trend was not apparent. Boice et al. (2006a) summarized this issue for the external radiation studies of nuclear workers. A radiation exposure association for any cancer site was only shown in studies that included relatively high dose levels as in the combined country studies (Cardis et al. 1995Cardis et al. , 2005), and studies of the Sellafield (Douglas et al. 1994) and Mayak ("
    [Show abstract] [Hide abstract] ABSTRACT: Workers involved in the nuclear fuel cycle have a potential for internal exposure to uranium. The present review of epidemiological studies of these workers aims to elucidate the relationship between occupational internal uranium exposure and cancer risk. Eighteen cohort and 5 nested case-control studies published since 1980 are reviewed. Workers occupationally exposed to uranium appear to be at increased risk of mortality from neoplasms of the lung, larynx, and lymphatic and haematopoietic tissue. Currently available evidence for a positive association between internal exposure to uranium and the risk of cancer is limited. The common weaknesses in reviewed studies include low statistical power and inaccurate assessment of internal exposure to uranium. Further investigations should focus on precise assessment of occupational exposure and address the issue of potential confounders.
    Full-text · Article · Feb 2008
Show more