Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
i The Participant Cohort The core of this report is a comparison of the mortality experience of nu- clear test participants and a comparable referent group of nonparticipants. This section contains a description of the participant cohort selection process. The participant cohort includes all military personnel identified by February 28, 1997, by the Defense Threat Reduction Agency (DTRA) as participants in at least one of the selected five series of U.S. atmospheric nuclear weapons tests. This study includes active duty personnel but does not include Reserve, National Guard, and Coast Guard personnel. The five test series Operations GREENHOUSE, UPSHOT-KNOTHOLE, CASTLE, REDWING, and PLUMB- BOB are the same series that were examined in the 1985 Medical Follow-up Agency (MFUA) study.~ As described earlier, these five series (consisting of 62 tests) were originally chosen for study from the 19 U.S. atmospheric nuclear weapons test series. Their selection was based on the availability and quality of records for personnel identification and radiation dosimetry and a design based on comparable numbers of participants at tests conducted in the Pacific and the continental United States. DTRA used the congressionally mandated and Department of Veterans Af- fairs-issued regulatory definition of participant: (~1) any U.S. military personnel who were present at the test site or who performed official military duties in connection with ships, aircraft, or other equipment in direct support of an at- mospheric nuclear test during its official operational period; (2) any U.S. mili- tary personnel who were present at the test site or other test staging area to per- form official military duties in connection with completion of projects related to the nuclear test, including decontamination of equipment used for the test, dur study. iSee Chapter 1 for a discussion of the 1985 publication and the rationale for the new 26
THE PARTICIPANT COHORT 27 ing the 6 months following the official period of an atmospheric nuclear test; or (3) any U.S. military personnel who served as members of the garrison or maintenance forces on Enewetak at any time from June 21, 1951, through July 1, 1952, after Operation GREENHOUSE; or from August 7, 1956, through August 7, 1957, after Operation REDWING (CFR, 1998a). Personnel in the last group, although not fitting the standard definition of test participation, were in- cluded by Congress and VA regulation as if they were participants because GREENHOUSE Shot ITEM and REDWING Shot TEWA, fired at Bikini Atoll, resulted in fallout on the Enewetak base camp, causing radiation exposure among DoD personnel who remained in the camp (Gladeck and Johnson, 1996; JAYCOR, 1995). Table 5-1 displays the operational period and the 6-month post-operational period for each of the five series. Identifying test participants was a difficult task, however, because a com- plete roster of test participants did not exist and the permanent DoD records of atmospheric tests did not contain the necessary identification information. Therefore, NTPR conducted large-scale searches of historical records, ranging from federal archives and records centers to private collections (Gladeck and Johnson, 1996~. For example, the Navy NTPR procedure for identifying partici- pants was first to identify the participating ships and squadrons through avail- able historical records. Deck logs, along with muster rolls and daily diaries, were then located to identify individual participants. For Army and Air Force NTPR teams, morning reports and personnel rosters of the units were located. Another source of information for the NTPR program has been a nation- wide toll-free call-in program set up by DNA for veterans of the atmospheric nuclear tests to report details of their participation in any test. When a call is received from a veteran or veteran's representative, an NTPR interviewer asks a standard set of questions and files the information. If a review of available rec- ords confirms the veteran's participation in the nuclear test series, the verified information is added to a computerized file that contains the participant data obtained primarily through record reviews. It is the latter, record-based file, rather than the initial direct contact, that is the source of data on participant identification for the current study. When the 1985 study began, individual NTPR teams were still rapidly identifying test participants. Since 1987, when DNA consolidated the service NTPR teams into a single operation, identification of new participants has slowed, but continues; participating units continue to be identified using newly discovered historical records, and the inclusion criteria for classifying partici- pants have been broadened.
28 U. . - V) ·_' a, o Cal o ._ o a, ho U. o - Ct o sit o ._ C) o In Cal Lo U. sit o C) ~ C) Cal O ~ a., ~ ~ 3 C,, ._ - ~ o o ._ Ct sit it · sit o PA ~ o ._ Ct o Cal ._ C) o 4~ Cal I o ._ Cal o He Z Cal ~ ¢ _ ,_ -_ ^ ~ Do o sit sit `` ~ ~ ;, ~ ~ ~ ~ O a.' ~ ~ ~ Z ~ ¢ so To ~ rat a' En ~ ~4 en be 4 - ¢ o ~cr en - ~ =- ¢ ~ ~ ~ ~ ~ ~ -9 'e ~c) ~'e Z ° E- 3 :~ a CQ ~ ~ Ed v ~ ~ Ox
THE PARTICIPANT COHORT TABLE 5-2. Estimates and Determined Extent of Participant Misclassification in the 1985 Dataset Assessed Misclassification GAOa oTAb1999 ReportC 1985 Report Total 46,186 46,18649,148 Wrongly included 14,854 4,5008,877 Wrongly omitted 28,215 15,00024,161 "Correct" total 59,547 56,68664,432 Additions due to decision 3,736 change, not error Total 68,168 aGeneral Accounting Office (1992~. bIJ.S. Congress Office of Technology Assessment (Gelband, 19924. Medical Follow-up Agency use of rosters supplied by the Nuclear Test Personnel Review Program as of 1997. The data file for 1985 included 49,148 records, 2,962 of which were excluded due to problem data. 29 From December 1993 through March 1997, DTRA transmitted to MFUA progressively updated data tapes that identified participants in the five series. Because MFUA was revisiting the questions first considered in the 1985 study, primarily because of inaccuracies in the DTRA-provided participant roster used in the 1985 analysis, both MFUA and DTRA provided intense and ongoing scrutiny of roster identification for the current study.2 RELATIONSHIP OF PARTICIPANT ROSTERS USED IN THE 1985 PUBLICATION AND THIS REPORT In the early 1 990s, DTRA (then the Defense Nuclear Agency) announced that the personnel dataset it had provided MFUA contained substantial errors of inclu- sion and exclusion. Because this dataset was the basis of MFUA's Five Series Study published in 1985, the U.S. General Accounting Office, the congressional Office of Technology Assessment, members of the U.S. Senate and House of Rep- resentatives, and MFUA itself recommended redoing the mortality analyses using a corrected dataset. MFUA further enhanced the study design (partially in re- sponse to criticism ofthe 1985 report) to include a military comparison cohort. Table 5-2 displays the extent of overlap between the participant cohort used for the 1985 publication and the cohort on which this current study is based. Eighty-four percent of the 1985 cohort is included in the current list. However, these people comprise only 57 percent of the current list. If we were to exclude 2Appendix D reviews the work done to validate participant cohort membership.
30 THE FIVE SERIES STUDY from this calculation the 3,736 personnel included in the current list solely be- cause of their presence in post-series rosters reflecting a post-1985 change in inclusion criteria rather than identification errors we still see that 60 percent of the current cohort was in the 1985 cohort. PARTICIPATION IN SERIES OTHER THAN THE SELECTION SERIES Participants were chosen for this study if they were assigned to military units that participated in at least one of the five selected series. The selection series is the first, chronologically, of the five series to which a member of the participant cohort could have been assigned. Three percent of this cohort par- ticipated in more than one of the five series and some participated in series other than the five. Table 5-3 illustrates, for the participant cohort, by selection series, the distribution of other test participation, according to the NTPR database. The count includes the five series studied in this report, any of the other 14 test se- ries, and assignment to Hiroshima or Nagasaki.3 3Although Hiroshima and Nagasaki were not test series, individuals who were as- signed to units in these areas in time proximate to the atomic bomb detonations are in- cluded in the NTPR database.
31 cd · · I) · - v) o · - c) - v) c) cd o I) a' lo .- cc · - o ce o In EN := 'e - ¢ If a' · An a o c) a of ~ ho ~ ~ . . . . . ~ ~ ~ o o oo -' o z m o m m v E~ ~Q ¢ v t:C ~Q o z C~ . ~ V: o . C~ ._ O . ~ O Ce ~OOO0 1 100 .... .. OOOOO OO o 00 00 V) V) O O ~ ~ ~ ~ _ 00 ~ ~ ~ 00 o 00 O ~. ~ O ~ ~4 ~ oo ~ ~ oo ~ ~ o ~ ~ o ~ ~ ~ ~ oo oo ~ ~ ~ ~ ~ ~ ~ ~ - oo ~ ~ c~ oo - oo ~oo ~oo o ~o oo ~ r~ ~cn ~u ~u: v ~c~ O ~a' ~a' ao · s ~· ~· s ~· 50 ~ 50 · 50 ~ s ~· s ~· s-- ·- ·- · s~ _ U) CQ U ~C ~C ~CO U ~U ~U: U ~U) ~S~' 5~' ~a) tV ~v a. ~a. ~ o o o o o o o o o o o o o C ~00 ~N O ~ c) + + + + + + + + + ~'e + + + o vo ~
