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 3
Long-Term Health Effects of Participation in Project SHAD (Shipboard Hazard and Defense) 1 Study Rationale and Overview The goal of our study was to determine the current health of participants in the Project SHAD (Shipboard Hazard and Defense) tests and compare their health with that of a comparable group of nonparticipant veterans from the same era. As a secondary goal, we hoped to be able to derive separate estimates of health effects for different kinds of participation, extending, if possible, to the estimation of separate effects for different agents used in Project SHAD. Data on current health status came primarily from a health survey, but mortality data were also collected and analyzed. A panel of expert advisors advised the Medical Follow-up Agency (MFUA) investigators in the conduct of this study. The contract for the study began on September 30, 2002, and ended on March 31, 2007. ORIGIN AND BACKGROUND Project SHAD was a series of tests conducted by the Department of Defense (DoD) in the 1960s to investigate the effectiveness of shipboard detection of and protection procedures against chemical and biological warfare agents (DoD, 2006). Within each test there were typically several separate trials involving exposure of vessels with various agents. In some cases, all the trials within a particular test used the same agent, but for some tests, different agents were used in different trials. Agents included chemical warfare agents sarin and VX; biological warfare agents Pasteurella tularensis,Coxiella burnetti, and staphylococcal enterotoxin B; chemical warfare simulants such as zinc cadmium sulfide; and biological warfare simulants such as Bacillus globigii and Serratia marcescens (see Appendix A for a complete list). Although the tests were originally classified, public and media interest has led the DoD to investigate these tests and to declassify and make publicly available relevant information from them. Further material describing the nature and conduct of the tests may be found on the study website (IOM, 2006) under “SHAD March meeting agenda.” Expert Panel and Meetings A panel of outside experts was appointed to advise us on the conduct of the study (see front matter). The panel members had expertise in epidemiology, biostatistics, study design and analysis, environmental epidemiology, infectious disease, and toxicology. The panel, appointed in accordance with standard policies of the National Academies, advised the study staff on the soundness of the proposed study design, monitored the conduct of the study, and reviewed the analyses and report of study findings. Because the panel was solely advisory to study
OCR for page 4
Long-Term Health Effects of Participation in Project SHAD (Shipboard Hazard and Defense) staff and did not provide advice to the federal government, its activities were not subject to the Federal Advisory Committee Act (FACA). Work Plan Assembling the Participant and Control Cohorts For each Project SHAD test, the DoD declassified and made public the following information: test name, military units involved, vessels involved, location, dates, and agents used. It is important to note that the declassified information gives only dates of the test and agents used, not the date of the individual trials and the agent(s) used at each trial. The DoD also assembled a list of individuals who were assigned to a test, whether or not they actually participated in any of the exposure trials within that test, and provided that list to the Department of Veterans Affairs (VA) and to us. We further refined these crew lists by abstracting and computerizing daily military unit records. We also asked the DoD to declassify and provide us the information necessary to determine trial participation. We compared the health of Project SHAD veterans to that of a group of comparable nonparticipants. For each ship in Project SHAD, with the exception of the light tugs, the DoD provided us with potential control ships of similar type. We did additional research on these potential control ships and then selected a nonparticipating ship of the same type and manning, with contemporary service, and obtained its personnel rosters. A roster of nonparticipant controls was then produced by a process similar to that for participants. Tracing Subjects Because our primary health outcome data were collected by health survey, we first had to locate the study subjects and obtain their current addresses and telephone numbers. This tracing involved both the use of commercial tracing firms and the National Institute for Occupational Safety and Health (NIOSH) access to address information from Internal Revenue Service (IRS) files. To make use of these sources of relatively up-to-date addresses, however, correct Social Security numbers (SSNs) are needed to match files; SSN was not used as the military service number until the 1970s. After conducting a pilot study, we decided to obtain SSNs from the individual hard-copy military service records of both participants and nonparticipants to supplement the SSNs obtained from readily available computerized sources, such as Beneficiary Identification and Records Locator Subsystem (BIRLS) (see Chapter 5 for further detail). Mortality Follow-Up The BIRLS search that was done to obtain SSNs also provided a date of death, which was used to screen out decedents before the interviewing process began. We followed the BIRLS search by a search of the National Death Index (NDI) files (see Chapter 4) to identify deaths and to obtain cause of death information. When we had identified a death using the BIRLS file, we obtained cause of death data from the NDI if the death occurred in 1979 or later. For deaths prior to 1979, the VA agreed to provide causes of death using records available in VA claims folders for participants only. Morbidity Follow-Up Using VA Records We matched the participant and nonparticipant rosters against the VA’s computerized files of inpatient (patient treatment file [PTF]) and outpatient (OPT) medical care. However, because veterans typically obtain only a small portion of their health care through the VA, the VA computerized morbidity data we obtained was useful only to compare survey respondents and nonrespondents and to validate self-reported morbidity information. Unfortunately, there was insufficient time for the latter analysis to be done.
OCR for page 5
Long-Term Health Effects of Participation in Project SHAD (Shipboard Hazard and Defense) Questionnaire Development We used a health survey to determine the general health of participants and nonparticipants as well as to identify possible adverse health effects (should they exist) that might be expected following exposure to the agents used in Project SHAD. However, in many instances, little is known about the long-term health effects following a particular exposure, making it difficult to know exactly what conditions should be captured by the follow-up survey. For that reason, we concentrated on the assessment of general health and also asked about a variety of medical conditions. There were several motivations for this. First, general health status provided a more accurate overall picture of the long-term health of these veterans than any single outcome measures. Second, because we did not know precisely which health outcomes might be associated with Project SHAD exposures, we needed to screen for a broad array of medical conditions. The questionnaire items on general health were the SF-36 (Ware et al., 2000), and current medical conditions were based on items taken from the National Health Interview Survey (NHIS) (NCHS, 2005), distributed by the Centers for Disease Control and Prevention (CDC). Not only do we have experience with the NHIS items from our earlier Edgewood study, but also the use of the SF-36 and NHIS items had the additional advantage of providing national data against which to compare the survey responses. Regarding specific health outcomes that we might expect to be sequelae of Project SHAD exposures, we asked our expert panel to identify specific outcomes for inclusion in the survey. We also asked a local firm to briefly review the toxicological literature for each agent used in Project SHAD. The executive summaries of these reports are in Appendix A, and the full text of the reports are available on the study website (IOM, 2006). For the special case of anticholinesterase nerve agents (i.e., sarin and VX), we used some items developed for an earlier survey, including two subscales from the Neuropsychological Impairment Scale (NIS), a survey measure with high validity and test-retest stability, which measured memory and attention problems (O’Donnell et al., 1993). The Data Collection Process The primary sources of morbidity data were mail questionnaires and telephone interviews. After the mail questionnaire was developed and approved by our institutional review board (IRB), we subcontracted with SRBI, a company experienced in telephone interviewing, to modify the mail questionnaire for use in telephone interviewing and to administer the survey by telephone interview. We worked with the National Academies’ Office of Contracts and Grants to develop a request for proposal, solicited bids, reviewed the bids for technical merit, and selected SRBI as our subcontractor. All data collection was approved by the National Academies’ IRB. Further details of data collection are available in Chapter 7. Sample Size Estimates When we began the study, we did not know the exact number of Project SHAD participants, but assumed the number was close to 4,000. Thus we assumed that there were 4,000 participants and that we would enroll an equal number of nonparticipants. Further, we assumed a location percentage of 85 percent and a survey response of 75 percent among those located, yielding 2,550 participant responses and an equal number of nonparticipant responses. Applying a standard formula to calculate minimum detectable relative risk at a 2-sided significance level of 5 percent and with 80 percent power, we obtained the following results: for a condition with 1 percent prevalence, there was a minimum detectable relative risk of 1.97; for a 5 percent prevalence, a 1.40 relative risk; for a 10 percent prevalence, a 1.29 relative risk; and for a 20 percent prevalence, a 1.21 relative risk. Thus, we determined that we would have sufficient power to detect 2-fold differences in health conditions provided they have a prevalence of 1 percent or more. In the end, we obtained nearly 2,700 responses from roughly 4,400 participants and more than 2,400 responses from roughly 5,200 nonparticipant controls, figures that are generally in line with our original power calculations.
OCR for page 6
Long-Term Health Effects of Participation in Project SHAD (Shipboard Hazard and Defense) Analytic Methods Outcome data included both nominal measures (e.g., prevalence of medical conditions) as well as continuous measures (e.g., scale measures for attention deficit). We used chi-square and t-tests for crude comparisons of these two types of measures, and logistic regression and general linear models analysis for adjusted comparisons. The adjusted comparisons took into account factors such as age, race, and pay grade. Cause-specific mortality was analyzed using standard mortality ratios and proportional hazards analysis, which allowed adjustments for factors such as age and race. Further details are available in Chapter 8. REFERENCES DoD (Department of Defense). 2006. Project 112. http://deploymentlink.osd.mil/current_issues/shad/shad_intro.shtml (accessed November 28, 2006). IOM (Institute of Medicine). 2006. Long-term health effects of participation in Project SHAD. http://www.iom.edu/CMS/3795/4909.aspx (accessed November 28, 2006). NCHS (National Center for Health Statistics). 2005. NHIS questionnaire, sample adult, adult conditions. http://www.cdc.gov/nchs/about/major/nhis/quest_data_related_1997_forward.htm (accessed November 28, 2006). O’Donnell, W. E., C. B. DeSoto, and J. L. DeSoto. 1993. Validity and reliability of the revised Neuropsychological Impairment Scale (NIS). Journal of Clinical Psychology 49:372-382. Ware, J. E., M. Kosinski, and J. E. Dewey. 2000. How to score version 2 of the SF-36 health survey. Lincoln, RI: QualityMetric Incorporated.
Representative terms from entire chapter: