6
Referent Cohort

GENERAL SCHEME

We decided on a two-stage scheme for selecting nonparticipant controls. For study personnel on ships, we began first by selecting a matching control ship for each participant ship. We then obtained the two quarterly rosters closest to, and bracketing, the dates of the corresponding test. Using these two quarterly rosters, we compiled a roster of all Navy personnel on the control ship who became part of the referent population for the study. In contrast to the process for identifying participants, we did not obtain and key information from the daily personnel diaries for the control ships. Although up to five Army light tugs, manned by Navy personnel, participated in several Project SHAD (Shipboard Hazard and Defense) tests, their complete personnel rosters were never found by the Department of Defense (DoD) or by us. Because there were relatively few Navy participants on these light tugs and because we had more than sufficient controls selected from other ships in the same tests, we did not select specific control ships for the Army light tugs.

Because we did not sample individuals from the control ships but instead took all persons on a selected ship (i.e., a census), our selection of controls is not strictly speaking a cluster sample. There is, nonetheless, an unmeasured component of variability associated with the sampling of ships. To properly estimate this component of variability would have required a much larger sample of ships than we had.

The process for Marine control units was similar in that each Marine participant unit was matched with a control unit. Finally, because there were so few identified Army, Air Force, and Coast Guard participants and controls (N = 160), we omitted them from most of the analyses.

DETAILED INFORMATION

Stage 1:
Selecting Ships

DoD personnel provided us a list of potential control ships for each participant ship and test, choosing potential control ships of the same type and class. We then developed formal control ship selection criteria that were sent to our expert advisory panel and to members of the Vietnam Veterans of America’s (VVA’s) Project SHAD Task Force for comment. The final control ship selection criteria are shown in Table 6-1. In general terms, we selected control ships to be the same type and class as the corresponding participant ship. We further selected control ships



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 19
Long-Term Health Effects of Participation in Project SHAD (Shipboard Hazard and Defense) 6 Referent Cohort GENERAL SCHEME We decided on a two-stage scheme for selecting nonparticipant controls. For study personnel on ships, we began first by selecting a matching control ship for each participant ship. We then obtained the two quarterly rosters closest to, and bracketing, the dates of the corresponding test. Using these two quarterly rosters, we compiled a roster of all Navy personnel on the control ship who became part of the referent population for the study. In contrast to the process for identifying participants, we did not obtain and key information from the daily personnel diaries for the control ships. Although up to five Army light tugs, manned by Navy personnel, participated in several Project SHAD (Shipboard Hazard and Defense) tests, their complete personnel rosters were never found by the Department of Defense (DoD) or by us. Because there were relatively few Navy participants on these light tugs and because we had more than sufficient controls selected from other ships in the same tests, we did not select specific control ships for the Army light tugs. Because we did not sample individuals from the control ships but instead took all persons on a selected ship (i.e., a census), our selection of controls is not strictly speaking a cluster sample. There is, nonetheless, an unmeasured component of variability associated with the sampling of ships. To properly estimate this component of variability would have required a much larger sample of ships than we had. The process for Marine control units was similar in that each Marine participant unit was matched with a control unit. Finally, because there were so few identified Army, Air Force, and Coast Guard participants and controls (N = 160), we omitted them from most of the analyses. DETAILED INFORMATION Stage 1: Selecting Ships DoD personnel provided us a list of potential control ships for each participant ship and test, choosing potential control ships of the same type and class. We then developed formal control ship selection criteria that were sent to our expert advisory panel and to members of the Vietnam Veterans of America’s (VVA’s) Project SHAD Task Force for comment. The final control ship selection criteria are shown in Table 6-1. In general terms, we selected control ships to be the same type and class as the corresponding participant ship. We further selected control ships

OCR for page 19
Long-Term Health Effects of Participation in Project SHAD (Shipboard Hazard and Defense) TABLE 6-1 The Process of Selecting Control Ships Step Procedure Review the DoD control ships list. Verify similarity of ship type (the control ship should be of the same type and class as the test ship, or a similar type and class). Determine size of the complement (the complement of the control ship should be at least as large as that of the test ship, or larger). Determine operating area (the control ship should have operated in the same ocean area as the test ship [in most cases, the Pacific Ocean]). Determine operational status (the control ship should have been in an operational status during the test period, meaning it should not have been in an extended overhaul or dry dock status). Determine home port (the control ship should have had the same home port). Designate potential control ships. Remove from consideration ships that do not meet all the criteria (except for home port).a Rank candidates. Assign a rank to each potential control ship depending on the similarity of control ship to participant ship (ties are allowed). Select the control ship.b Take the control ship with the highest rank; if there are ties, select randomly. aDeviations from these criteria were avoided when possible, but could be necessitated by factors that prevailed at the time of the tests; e.g., there may have been no similar type ship in the DoD list of potential control ships, all DoD-listed potential control ships may have been in a nonoperational status during the exact period of a test, or the potential control ships may not have had the same home port as the test ship. bA test ship that served in multiple tests may have more than one control ship selected because of changes in a potential control ship’s fulfillment of one or more of the above criteria; e.g., it may have been operational during the period of one test, but in dry dock during another test. with a complement, operating area, and home port similar to that of a Project SHAD vessel. When there were multiple possible control ships, we ranked them in order of desirability and selected the closest match. To assist in characterizing potential control ships, we hired Jim Quinn, Commander, USN retired, as a consultant. For the few Marine units, we selected a similar unit in operation at the same time. In general, we tried to select the identical unit in a parallel battalion or division. The final list of control units is shown in Table 6-2. Stage 2: Selecting Individual Subjects Once a control ship had been selected, we used a similar process as was used for participant ships; that is, we obtained the quarterly BuPers reports for the corresponding time periods. However, we did not make use of personnel diaries. We keyed the entries from unit records and produced a list of control subjects, identified by name, military service number, and rate or job title (for officers). The process for Marine control units was similar to that for Marine participants, except that, again, we did not use personnel diaries. In some cases, we selected Marine nonparticipants from the same unit as the participants. We do not think that omitting potential controls who would have been identified solely from personnel diaries is a substantial omission; only 5 percent of Navy participants were identified solely from personnel diaries. In preparing a participant roster for their own purposes, the DoD supplemented the unit record information on participants with other data from other sources, for example, Social Security numbers (SSNs) from individual personnel records, addresses from other sources, and so on. Because the DoD did not assemble a control roster, we had to begin ab initio from unit records to assemble the control roster for our study, and our controls never had initial identifying information beyond name, service number, and rate. Although we undertook a similar process as the DoD, we began our identification of controls later, and thus our control subjects were typically less well identified than participants and harder to trace, locate, and contact.

OCR for page 19
Long-Term Health Effects of Participation in Project SHAD (Shipboard Hazard and Defense) TABLE 6-2 List of Participant and Control Units Showing Selected Characteristics Participant Unit Control Units Test Name and Unit Name Ship Type Operating Area Unit Name Ship Type Operating Area Autumn Gold USS Granville S. Hall (YAG-40) Auxiliary ship Pearl Harbor USS Interceptor (AGR-8) Radar picket ship Picket Station 1 (Canada), Treasure Island, Portland USS Navarro (APA-215) Attack transport Pearl Harbor USS Talladega (APA-208) Attack transport Long Beach USS Tioga County (LST-1158) Tank landing ship Pearl Harbor USS Vernon County (LST-1161) Tank landing ship Japan, Philippines USS Carpenter (DD-825) Destroyer Pearl Harbor USS John R. Craig (DD-885) Destroyer San Diego USS Hoel (DDG-13) Guided missile destroyer Pearl Harbor USS Towers (DDG-9) Guided missile destroyer San Diego VMA 214, Marine Air Group 13 — — VMA 332, Marine Air Group 14 — — Big Tom USS Granville S. Hall (YAG-40) Auxiliary ship Pearl Harbor USS Oxford (AGTR-1) Auxiliary ship Subic Bay, Philippines USS Carbonero (SS-337) Submarine — USS Raton (SS-270) Submarine San Clemente, San Diego Copper Head USS Power (DD-839) Destroyer — USS Gyatt (DD-712) Destroyer Norfolk, Portsmouth DTC Test 69-10 USS Fort Snelling (LSD-30) Dock landing ship — USS Spiegel Grove (LSD-32) Dock landing ship Little Creek, VA; Onslow Beach, SC; Morehead City, NC; Vieques Landing Force Carib 1-69/BLT 1/8 — — 1st Battalion, 6th Marines, 2nd Marine Division — — VMA 324, MAG-32 — — VMA 324, MAG-32 — — DTC Test 69-31 USS Herbert J. Thomas (DD-833) Destroyer Pearl Harbor USS Agerholm (DD-826) Destroyer San Diego DTC Test 69-32 USS Granville S. Hall (YAG-40) Auxiliary ship Pearl Harbor USS Jamestown (AGTR-3) Auxiliary ship South China Sea, Thailand, Vietnam, Subic Bay, Special Operations Eager Belle I USS George Eastman (YAG-39) Auxiliary ship Pearl Harbor and maneuvers USS Interceptor (AGR-8) Radar picket ship Picket Station 1, Picket Station 9, San Francisco Eager Belle II USS George Eastman (YAG-39) Auxiliary ship Pearl Harbor USS Interceptor (AGR-8) Radar picket ship San Francisco Picket Stations 1, 3, 9

OCR for page 19
Long-Term Health Effects of Participation in Project SHAD (Shipboard Hazard and Defense) Participant Unit Control Units Test Name and Unit Name Ship Type Operating Area Unit Name Ship Type Operating Area USS Tioga County (LST-1185) Tank landing ship Pearl Harbor USS Vernon County (LST-1161) Tank landing ship Yokosuka, Kobe, Taiwan, Po Hong Do, Okinawa USS Carpenter (DD-825) Destroyer Pearl Harbor USS Agerholm (DD-826) Destroyer Subic Bay, Hong Kong, Manila, Yokosuka, Okinawa USS Granville S. Hall (YAG-40) Auxiliary ship — USS Interdictor (AGR-13) Radar picket ship — USS Navarro (APA-215) Attack transport — USS Noble (APA-218) Attack transport — Marine Medium Helicopter Squadron 161 — — Marine Medium Helicopter Squadron 161 — — Errand Boy USS George Eastman (YAG-39) Auxiliary ship Pearl Harbor USS Interceptor (AGR-8) Radar picket ship San Francisco, Picket Station 7 Fearless Johnny USS George Eastman (YAG-39) Auxiliary ship — USS Oxford (AGTR-1) Auxiliary ship Hong Kong, Subic Bay USS Granville S. Hall (YAG-40) Auxiliary ship Pearl Harbor USS Georgetown (AGTR-2) Auxiliary ship Hong Kong, Subic Bay Flower Drum I USS George Eastman (YAG-39) Auxiliary ship Pearl Harbor USS Interceptor (AGR-8) Radar picket ship San Francisco, San Diego, Radar Picket Stations 1, 5, 7 USS Granville S. Hall (YAG-40) Auxiliary ship — USS Interdictor (AGR-13) Radar picket ship — Folded Arrow USS Granville S. Hall (YAG-40) Auxiliary Ship Pearl Harbor USS Oxford (AGTR-1) Auxiliary ship Subic Bay, South China Sea USS Carbonero (SS-337) Submarine Pearl Harbor USS Tunny (SS-282) Submarine Subic Bay, Special Ops Half Note USS George Eastman (YAG-39) Auxiliary Ship — USS Oxford (AGTR-1) Auxiliary ship Taiwan, Hong Kong, Special Operations USS Granville S. Hall (YAG-40) Auxiliary Ship Pearl Harbor USS Jamestown (AGTR-3) Auxiliary ship Malaysia, Taiwan, Special Operations USS Carbonero (SS-337) Submarine — USS Tunny (SS-282) Submarine Pearl Harbor, Subic Bay Light tug 2085 — — None selected — —

OCR for page 19
Long-Term Health Effects of Participation in Project SHAD (Shipboard Hazard and Defense) Participant Unit Control Units Test Name and Unit Name Ship Type Operating Area Unit Name Ship Type Operating Area High Low   —       USS Berkeley (DDG-15) Destroyer   USS Lynde McCormick (DDG-8) Destroyer Hong Kong, Yokosuka, San Diego USS Fechteler (DD-870/DDR-870) Destroyer — USS John R. Craig (DD-885) Destroyer San Diego USS Okanogan (APA-220) Attack transport — USS Montrose (APA-212) Attack transport San Diego, Pearl Harbor, San Clemente USS Wexford County (LST-1168) Tank landing ship — USS Washoe County (LST-1165) Tank landing ship Numazu, Yokosuka, Okinawa Magic Sword USS George Eastman (YAG-39) Auxiliary ship Pearl Harbor USS Interceptor (AGR-8) Auxiliary ship Durban, Subic Bay Purple Sage   —       USS Herbert J. Thomas (DD-833) Destroyer   USS Agerholm (DD-826) Destroyer San Diego Scarlet Sage   —       USS Herbert J. Thomas (DD-833) Destroyer   USS Agerholm (DD-826) Destroyer San Diego, Long Beach Shady Grove USS Granville S. Hall (YAG-40) Auxiliary ship Pearl Harbor USS Interceptor (AGR-8) Auxiliary ship Bremerton, Wash; Panama Canal Zone; Guantanamo Bay, Cuba; Portsmouth, VA VMA 214, MAG13 — — VMA214, MAG 13 — — MWSG 13, Marine Air Group 13* — — MWSG, MAG 13 — — MABS 13, Marine Air Group 13* — — MABS 13, MAG 13 — — HMM 161, Marine Air Group 13 — — HMM 161, MAG 13 — — Light tugs 2080, 2081, 2085, 2086, 2087 — — None selected — — Speckled Start USS Granville S. Hall (YAG-40) Auxiliary ship Pearl Harbor USS Oxford (ATGR-1) Auxiliary ship Subic Bay, South China Sea *Original participant files contained personnel in these units who were subsequently removed from the study.