4
Records-Based Data Sources
INTRODUCTION
In this chapter, the records-based data sources are briefly discussed, while a separate chapter (Chapter 7) describes the health survey. Records-based sources include the fact sheets published by the Department of Defense (DoD) that identified the Project SHAD (Shipboard Hazard and Defense) tests and described the military units of the participants in each test (DoD, 2006). A database assembled by the DoD contained records identifying each Project SHAD participant, with additional information from Department of Veterans Affairs (VA) records, such as Social Security number (SSN) or address. We received these records and further processed them to include data from the Defense Manpower Data Center, the VA’s Beneficiary Identifier and Records Locator Subsystem (BIRLS), and the National Death Index (NDI), among other sources. We also obtained data from hard-copy sources, such as the individual’s military personnel record, and included these data in the study’s master file. A parallel effort assembled similar data for military veterans who were not participants in Project SHAD. This chapter will briefly describe the data sources used to assemble the study’s data files. More detail on VA records resources is available in Boyko et al., 2000.
DATA SOURCES USED TO IDENTIFY PARTICIPANTS AND CONTROLS
DoD Fact Sheets
Basic information about the Project SHAD tests came from fact sheets prepared by the DoD and posted on a DoD website (DoD, 2006). These were updated as additional information came to light about the tests. The DoD website contains a list of all Project SHAD tests, together with links to the fact sheet for each Project SHAD test. The fact sheets for each Project SHAD test give information on the dates of the test, military units involved, agents used in the test, and so on. The listings of military units involved in each test provided the starting point for our research using military unit records.
Project SHAD Technical Reports
Although Project SHAD technical reports remain, in general, classified, we were sent selected declassified sections from some final reports. Of particular use were Tables 12 through 15 of Volume II of the final report for DTC Test 69-10, which contained estimates of contamination for certain individual participants. See Chapter 8 for further details.
Military Unit Records
DoD personnel assembled the initial roster of Project SHAD participants using military unit rosters, which we also consulted. For Navy personnel, the quarterly unit rosters for enlisted personnel, BuPers Report 1080-14, record every enlisted person on that ship on the given day that ends a quarter (e.g., March 31), showing name, service number, and rate (e.g., machinist mate); the listing is arranged by rate. Officers present on the ship on the given date of a quarterly report are listed separately on the Officer Distribution Control Report, which shows name, service number, and job title (e.g., commanding officer). We also obtained and reviewed the daily personnel diaries for each ship in each Project SHAD test, as well as for control ships. The daily personnel diaries list individuals who have come on or left the ship, along with a description of the reason for the movement on or off ship (e.g., absent of sailing). Marine participants were occasionally listed in Navy unit records, particularly the daily personnel diaries.
The Marine unit records are similar to those of the Navy. The monthly personnel roster is a list of Marines by name, military service number, and pay grade. We used these rosters as well as the company diaries, which document the movement of individuals, to assemble the Marine participant and control cohorts.
DATA SOURCES USED TO LOCATE AND FOLLOW-UP PARTICIPANTS AND CONTROLS
BIRLS
The VA’s BIRLS file is a computer file that identifies beneficiaries and locates their VA claims records. We used BIRLS records as the first step in our mortality follow-up process, as this file contains the date of death for deceased veterans. A BIRLS record may also contain a military service number as well as an SSN, making it a potential cross-index of service numbers and SSNs. Although our primary method of obtaining BIRLS data was by matching a computer file with many records (typically tens of thousands), we also searched the BIRLS file by individual record, using TARGET access.
Registry and Individual Service Records
Each military veteran has an individual personnel folder, which contains, among other things, identifier and demographic data such as name, rank, military service number, and SSN. These records are housed at the National Personnel Records Center in St. Louis, Missouri, and are indexed by a computerized registry file. Access to individual military records was granted by the individual service branch that owned the records.
MSN/SSN File and Bidex/Tridex
When military service numbers (MSN) were replaced by SSNs as the military services’ identification number starting in July 1969, a number of MSN/SSN cross-index files were created. The MSN/SSN file is such a file and contains several million records with name, MSN, and SSN. We used the MSN/SSN file to try to obtain SSNs for veterans for whom we had only an MSN. The Bidex and Tridex files are cross-index files with MSN and SSN and name, MSN, and SSN, respectively. However, both the computerized MSN/SSN file and the Bidex and Tridex files are only partial cross-indices, for reasons unknown to us.
National Death Index
The NDI is a computer file maintained by the National Center for Health Statistics. We used the NDI both to identify decedents and to provide causes of death. Because the NDI contains death information from 1979 on, other data sources must be used to obtain fact and cause of death prior to 1979.
Commercial Address Tracing Firms
A number of firms can obtain a current address by matching against their files using name and SSN. We made use of Intellius and Choice Point in this study.
The National Institute of Occupational Safety and Health and the IRS
Special legislative authority exists for the director of the National Institute of Occupational Safety and Health (NIOSH) to request mailing addresses from the Internal Revenue Service (IRS) to locate individuals who “may have been exposed to occupational hazards during active military, naval, or air service …” (Public Law 96-128, section 502). We used NIOSH/IRS addresses in some of our attempts to contact study subjects.
REFERENCES
Boyko, E. J., T. D. Koepsell, J. M. Gaziano, R. D. Horner, J. R. Feussner. 2000. US Department of Veterans Affairs medical care system as a resource to epidemiologists. American Journal of Epidemiology 151(3):307-314.
DoD (Department of Defense). 2006. Project 112. http://deploymentlink.osd.mil/current_issues/shad/shad_intro.shtml (accessed November 28, 2006).