1 15 4 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 054 Folder The folder containing the original item. 1441 Series The series number of the original item. Series III Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Newman, Don M. Title A name given to the resource Letter: from Don M. Newman to Alan Cranston, 3 November 1987 Subject The topic of the resource Agent Orange Exposure Study Vietnam Experience Study Selected Cancers Study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/5e96f5eebbc8e3cd125eb07257433b1e.pdf 78b3ef75aa9c59c169a1dc6ca42549b1 PDF Text Text Item ID Number 01728 Author Centers for Disease Control, Public Health Service, U.S. Ruport/Articlo Tltlfl Protocol for Epidemiologic Studies of the Health of Vietnam Veterans Journal/Book Title Year MOUth/Day Color November n Number of Imaoes 105 Descrlpton Notes Monday, June 11, 2001 Page 1729 of 1793 �PBOTOCOL FOR EPIBEMIOLQGIC STUDIES OF THE HEALTH OF VIETNAM VETERANS 1. Cohort Study of the Long-Term Health Effects of Exposure to Agent Orange In Vietnam, 2. Cohort Study of the Long-Term Health Effects of Military Service in Vietnam, and 3. Case-Control Study to Determine the Risks for Selected Cancers Among Vietnam Veterans. to be conducted by CENTERS FOR DISEASE CONTROL PUBLIC HEALTH SERVICE U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES Atlanta, Georgia 30333 November 1983 U.S. Department of Health and Human Services Public Health Service Centers for Disease Control �NOTICE At the time this document was printed (December, 1983) these protocols had not yet received approval for protection of human subjects from the Institutional Review Board of the Centers for Disease Control, nor clearance by the Office of Management and Budget. Both approvals are required for implementation of the studies described in this document. �Contents CENTERS FOR DISEASE CONTROL Protocol for Epidemiologic Study of the Health of Vietnam Veterans 1. Introduction 2. Background 2.1. 2.2. 2.3. 2.4. 2.5. Herbicide Usage in Vietnam Health Effects of Herbicides and Dioxin Diseases Affecting U.S. Troops in Vietnam Current Health of Vietnam Veterans Long-Term Health Status of Servicemen and Veterans 3. Study Design Overview 3.1. Agent Orange Study 3.2. Vietnam Experience Study 3.3. Selected Cancers Case-Control Study 4. Study Procedures 4.1. Selection of Study Subjects 4.1.1. Agent Orange Study 4.1.2. Vietnam Experience Study 4.1.3. Selected Cancers Case-Control Study 4.2. Location of Study Subjects 4.3. Ascertainment of Health and Exposure Status 4.3.1. Agent Orange and Vietnam Experience Studies 4.3.1.1. Mortality Information 4.3.1.2. Morbidity and Exposure Information A. Health Interviews —Sociodemographic Information —Medical History —Environmental and Occupational Exposure Information --Military History B. Medical and Psychological Examinations; Laboratory Tests 4.3.2. Selected Cancers Case-Control Study —Sociodemographic Information —Family History of Cancer —Medical History —Environmental and Occupational Exposure Information —Military History 4.4. Sample Sizes, Statistical Power, and Participation Rates 4.4.1. Agent Orange and Vietnam Experience Studies 4.4.2. Selected Cancers Case-Control Study Page 1 3 3 4 5 5 5 6 6 8 8 10 10 10 15 16 18 19 20 20 20 21 21 22 22 22 23 24 25 25 25 25 25 26 26 28 �4.5. Pretests and Pilot Studies 4.5.1. Agent Orange and Vietnam Experience Studies 4.5.1.1. Military Records Pretests 4.5.1.2. Location Rate, Participation Rate, and Instrument Assessments 4.5.2. Selected Cancers Case-Control Study 4.6. Data Analysis and Quality Control 4.6.1. Timing of Analyses 4.6.2. Summary of Analytical Approach 4.6.3. Quality Control Contents Page 29 29 30 31 31 32 32 33 35 5. Inferences from Possible Study Findings; Study Limitations 37 6. Report of Study Findings 39 7. Timetable, Milestones, and Reports 40 8. Investigators 41 9. Protocol Review; Study Oversight 42 10. Tables Table 1, Table 2. Table 3. Table 4. Table 5. 11. Appendices Appendix Appendix Appendix Cumulative Expected Numbers of Deaths by Cause in a Hypothetical Cohort of 6,000 Men Aged 22 in 1968 and Followed Through 1984 (17 Years) Power To Detect Various Relative Risks in the Agent Orange and Vietnam Experience Studies, by Prevalence of Condition in "Unexposed" Group Selected Health Outcomes Reported To Be Associated with Exposure to TCDD—Animal and Human Literature Estimated Prevalence of Vietnam Service and Expected Number of Cases of Cancer for the Selected Cancers Case-Control Study in Males Aged 30-54 in 1986 in the SEER Areas Power of Selected Cancers Case-Control Study To Detect Increased Relative Risks Protocol Outline (November 1982) Literature Review Sample Selection Using Telephone Random Digit Dialing Appendix Topical List of Questionnaire Items for Agent Orange and Vietnam Experience Studies Appendix Topical List for Examination and Laboratory Testing for Agent Orange and Vietnam Experience Studies Appendix F. Topical List of Questionnaire Items for Selected Cancers Case-Control Study 12. References 43 44 46 47 48 50 63 78 79 81 83 85 �Page 1 ]. Introduction In response to the concerns of Vietnam veterans regarding their health, the Centers for Disease Control (CDC) herein proposes three distinct but related studies which are in addition to CDC's ongoing birth defects study. CDC believes that they provide the best opportunity to answer questions of importance to Vietnam veterans and their families, even though some aspects of the proposed studies are not scientifically ideal. The concerns of the nearly 3 million men who served in Vietnam for their health are real. If Vietnam veterans are at an increased risk of ill health, the personal and public health impact cannot be overestimated. In any case, the concerns and uncertainty alone represent a significant problem. CDC will be pleased to be able to provide a service to the nation *s Vietnam veterans by conducting these studies to evaluate their health. In this document CDC proposes two historical or retrospective cohort studies and one case-control study. One of the cohort studies will compare the health of a group of male U.S. Army veterans of the Vietnam conflict with the health of a group of male Army Vietnam-era veterans who did not serve in Vietnam. The purpose of this study will be to make an assessment of the possible health effects of the general Vietnam service experience, and will hereafter be referred to as the "Vietnam Experience" study. The other cohort study, which is designed to evaluate the health effects of possible exposure to herbicides (primarily Agent Orange), will compare the health of three groups or cohorts of male Vietnam veterans who differ in their probable level of exposure to Agent Orange and other herbicides. This second cohort study, to be referred to as the "Agent Orange" study, will also be limited to veterans of the Army. The third study will be a case-control study to evaluate the risk of contracting soft tissue sarcoma and lymphoma among Vietnam veterans (and/or those exposed to herbicides); this study will be designated as the "Sarcoma/Lymphoma" study. It is a critical part of CDC's effort because there is a specific concern about veterans' risk for these cancers, and the cohort studies are not large enough to provide answers about them. Cases and controls for the Sarcoma/Lymphoma study will be limited to males who were of draftable age during the Vietnam conflict, and will include veterans from all branches of the military. Each of the two cohort studies will have three major components: 1) a mortality assessment (mortality follow-up will be repeated every 5 years for the forseeable future); 2) a health interview; and 3) a clinical and laboratory assessment. The studies will have several other features in common. However, the sampling plans will differ and some of the health outcomes measured in the interviews and clinical assessments will receive different emphases in the two studies. The Sarcoma/Lymphoma case-control study will involve a health and exposure interview. laken together, the three studies proposed in these protocols, along with CDC's ongoing birth defects strudy, represent a fairly comprehensive approach to the health concerns of Vietnam veterans. In many respects, the studies are complementary to one another. Without conducting each of the three studies proposed herein, the CDC does not believe it can adequately assess the concerns of Vietnam veterans. �Page 2 This set of protocols presents the general framework of CDC's proposed studies. The studies will be very large and complex undertakings and not all details are presented; indeed, many details cannot be presented until work proposed in the protocols is done. CDC's policy of openness about its plans will continue as the studies progress. Historical Note on CDC's Involvement Public Law 96-151 requires that the Veterans Administration (VA) conduct an "epidemiological" study of U.S. veterans to assess the possible health effects of exposure to herbicides and dioxin during the Vietnam conflict. Public Law 97-72 expands this mandate to include the study of other environmental exposures which may have occurred in Vietnam. At about the time Public Law 96-151 was enacted, CDC proposed its ongoing birth defects study to assess the Vietnam veteran's risk of fathering children with congenital malformations. The responsibility for the design, conduct, and analysis of studies responsive to these laws was transferred, by an Interagency Agreement, from the VA to CDC in mid-January 1983. In November 1982 a team of CDC scientists prepared a "protocol outline" (Appendix A) which set down the rudiments of CDC's study plans, and the outline served as the basis for the Interagency Agreement* The present document expands on and supplements the ideas contained in the November 1982 "protocol outline." �Page 3 2. Background The review of background information regarding the possible health effects of military service in Vietnam presented here is intentionally very brief. It is intended to give an appreciation of the rationale for CDC's proposed studies. Those who desire more detail on health effects are referred to Appendix B and this document's reference list; the comprehensive review of the literature which was conducted for the VA is a particularly good source of information on herbicides. Those familiar with the literature can proceed directly to Section 3. 2.1 Herbicide Usage in Vietnam Herbicides were used for three principal purposes during the Vietnam war: defoliation - to cause trees and plants to lose their leaves in order to improve observation; crop destruction - to destroy the food value of certain crops; and, on a smaller scale, to clear vegetation around fire bases and other installations, around landing zones, and along lines of communication. The use of herbicides during the Vietnam war began in 1962, was greatly expanded during 1965-1966, and peaked from 1967-1969. In 1969 it was reported that mice exposed to certain herbicide components bore offspring with birth defects. Between 1970 and 1971 the use of herbicides was phased out in Vietnam. The tactical military project for the aerial spraying of herbicides in South Vietnam was named "Operation Ranch Hand;" this program used fixed-wing aircraft and disseminated the bulk of the herbicides used in Vietnam. Smaller quantities of herbicides were applied from helicopters, trucks, riverboats, and by hand applicators. At least two groups of U.S. personnel appear to have been at risk for exposure to herbicides—those involved in the transport and dissemination of the agents and those exposed at the time of spraying, such as troops on the ground. Although exposures may have occurred during transportation (e.g., because of damage to containers), aircraft crew — particularly flight mechanics and crew chiefs — were thought to be at greatest risk. Even though the major portion of herbicides used was disseminated by Ranch Hand, a significant and even major source of exposure of ground troops may have been from non-Ranch Hand applications. Records of Ranch Hand missions are contained on the so-called "Herbs" computer tapes, and records of other herbicide applications are on the "Services Herbs" tapes (see Section 4.1.1). Herbicides used for military purposes during the war were identified by color bands on their containers (e.g., orange, white, purple, etc.). The herbicide known as Agent Orange was most widely used in Vietnam. It was a 50:50 mixture by weight of the butyl esters of two phenoxy acid herbicides, 2,4-dichlorophenoxy acetic acid (2,4-D) and 2,4,5-trichlorophenoxy acetic acid (2,4,5-T). In addition, TCDD (2,3,7,8-tetrachlorodibenzo-para-dioxin, "dioxin") was a synthetic contaminant of 2,4,5-T; levels of TCDD contamination of Agent Orange ranged from 0.02 to 47 ppm, with a mean of about 2 ppm (Young et al., 1978). �Page 4 2.2. Health Effects of Herbicides and The herbicide contaminant TCDD is considered to be one of the most toxic compounds known. Thus, any interpretation of abnormal findings related to 2,4,5-T exposure must take into consideration the presence of varying or undetermined amounts of TCDD. Single oral TCDD LDSC's range from 0,6-2.0 ug/kg in the guinea pig to JJ57-5051 ug/kg in the hamster (Scbwetz et al., 1973; Olson et al., ]980; Kociba and Schwetz, J982). A wide variety of health effects have been observed following administration of TCDD to experimental animals. Acute and chronic toxic effects in animals include cercinogenesis, maternally mediated teratogenesis, hepatic necrosis, decreased body weight, alopecia, chloracne, thyrous atrophy, adrenal hemorrhage, immunosuppression (e.g., decreased cell-mediated immunity and lymphopenia), and other hematologic changes. In humans, toxic effects have been reported after occupational exposure during the industrial synthesis of 2,4,5-trlchlorophenol (TCP) and 2,4,5-T, after exposure in factories and in the surrounding environment following explosions which occurred during the synthesis of TCP, and after exposure to herbicides and other materials containing TCDD. Many of these studies had no, or Inadequate, controls; exposure was usually of unknown magnitude and duration, to what were often mixtures of chemicals; and the total number of exposed persons was usually not reported. Available data on dermatologic, hepatic, neuropsychologic, iromunologic, carcinogenic and reproductive effects are reviewed in Appendix £ and briefly summarized below. The most frequent and consistent acute health effect of TCDD exposure is chloracne, a refractory acne which is also caused by exposure to certain other halogenated hydrocarbons. Chloracne may be accompanied by hyperpigmentation and/or birsutlsm and can persist for many years after exposure. Porphyria, a liver disorder resulting in abnormalities of heme pigment metabolism and often accompanied by skin manifestations, has been reported after several industrial accidents. Other hepatic effects include structural alterations, changes in the biliary system and alterations in serum levels of certain liver enzymes. Neurological and/or psychological effects have been reported after most episodes of accidental industrial exposures. Common complaints have included Irritability, fatigue, weakness and pain, headaches, sexual dysfunction and loss of appetite. Signs of peripheral neuropathy, including decreased nerve conduction velocity have been reported. Immunological effects have been observed In experimental animals, including changes in thymus and other lymphoid tissues. TCDD also suppresses immune function, particularly thymic-dependent function. Reduced mitogen responsiveness, and impaired skin-graft rejection and delayed hypersensitivity responses have been observed in animal species, TCDD is carcinogenic in rats and mice; it appears to act as a tumor promoter in these species. Evidence is accumulating that human occupational exposures may be associated with an increased risk of soft tissue sarcoma and lymphoma. Somewhat weaker evidence suggests that herbicide exposure may be associated with nasal and nasopharyngeal cancers. Allegations that herbicide exposure is associated with primary liver cancer have emanataed from Vietnam. �Page 5 Reproductive effects in animals appear to be limited to maternal (fetal) exposures; the few studies that have addressed the possibility of paternally mediated effects have not shown differences in rates of poor reproductive outcomes between the exposed and non-exposed. The human data on reproductive outcomes after exposure is also generally negative, but most specific poor reproductive outcomes are rare, and the studies of men exposed in industrial settings have been relatively small. 2.3. Diseases Affecting U.S. Troops in Vietnam Overall, the average annual hospital admission rates for diseases among soldiers in Vietnam (351 per 1,000 per year) were about 30% lower than for the China-Burma-India and Southwest Pacific theaters in World War II and 40% lower than for the Korean war. Malaria was the most significant medical problem in Vietnam, accounting for the greatest number of lost man-days. Diarrheal, skin, and venereal diseases were also significant problems. Before 1968 neuropsychiatric disorders were not unusually frequent among men serving in Vietnam, but by 1970 they became the second leading disease problem. 2.4. Current Health of Vietnam Veterans Many Vietnam veterans believe that they may be at increased risk for a wide variety of diseases. Concerns voiced by Vietnam veterans include (to name just a few) dermatologic conditions, neurological disorders, reproductive problems, cancer, and infections. Unfortunately, little objective evidence is available regarding the health of Vietnam veterans relative to the health of other men of similar age. Indeed, this lack of data is a major reason for the studies proposed here. Data are available, however, for certain health-related issues such as psychosocial adjustment. Psychosocial adjustment problems could, in one sense, be considered health outcomes and, in another sense, causes or effects of other health outcomes. The literature suggests that Vietnam veterans differ from other veterans and from non-veterans in the level of their educational achievement, occupational status, psychological symptoms (especially anxiety, depression, and anger), drug and alcohol use, and frequency of arrest. 2.5. Long-Term Health Status of Servicemen and Veterans An additional literature review was done to provide background for the Vietnam Experience study. The most important finding of this review was not unexpected: because of medical selection at the time of induction into the military, ex-servicemen, especially officers, enjoy better long-term health than their counterparts who did not serve in the military. It was thought that one would find many reports of studies that compared the health of men who had seen combat with the health of contemporary men who had not participated in combat. CDC was unable, despite an extensive search, to find such reports. The details of CDC's search and a review of some of the reports found can be found in Appendix B. �Page 6 3. Study Design Overview The purpose of this section is to provide a summary of the general rationale for CDC's recommendation for three separate studies; this will be useful background for the subsequent description of the proposed study procedures. The section reiterates and, in some respects, amplifies the "protocol outline" prepared by CDC in November 1982 (Appendix A). 3.1. Agent Orange Study A good design for a historical cohort study of the possible health effects of herbicide exposure would involve the use of two groups of men who were as similar as possible except for their exposure to the herbicide. Ideally, one group would be free from all exposure, and the other would have been subjected to "meaningful" exposure. It appears that such an ideal is not attainable. Obstacles include 1) the fact that the military records that must be used to assess exposure were made during a war and are, therefore, of uneven quality; 2) the inability to define objectively "meaningful" exposure; 3) the difficulty in ensuring that veterans who were possibly or probably exposed (by whatever measure) are comparable (with respect to all things that might influence health) to veterans who were not exposed. These obstacles are formidable impediments to the accurate assessment of health effects of herbicide exposure. In view of these obstacles, CDC proposes what it considers the best (albeit imperfect) approach to studying this issue. The important records that give information about troops are the company morning reports and the battalion journal files. The morning reports can be used to document the presence or absence of individual servicemen on a daily basis, and the daily journal files will indicate the locations of companies in time and space. The major herbicide records are those that document the time and location of fixed-wing aircraft applications of herbicide (Ranch Hand missions), base perimeter applications records, and information about Ranch Hand mission aborts. The choice of an individual for inclusion in the "exposed" cohort will be based on a measure of company proximity in time and space to herbicide applications, as documented by these records. Members of the "non-exposed" cohort will likewise be selected according to a measure of their company's distance in time and space from any herbicide applications. Because of the uncertainties involved in assessing exposure, the two cohorts will hereafter be denoted by the terms "likely exposed" and "likely not exposed," respectively. The company records may contain gaps (i.e., whole periods of time missing) and are probably quite variable in terms of quality and detail, because they were created during the war. The herbicide usage records are known to contain errors with respect to the time and location of applications, and the degree of their completeness is unknown. They are far from ideal as the starting point for an historical cohort study. There may be opportunities to assess the accuracy and completeness of the herbicide usage records, and every effort will be made to pursue these opportunities (Section 4.1.1.). However, there are no possibilities for similar checking of the company troop records. Thus, the categorization of individuals with respect to their potential for herbicide exposure will be uncertain and will forever remain so. �Page 7 The desire to ensure that troops classified as "likely exposed" to herbicides are comparable to "likely not exposed" troops with respect to other factors that might influence health also makes it difficult to design an "ideal" study. The underlying problem is that the use of herbicide was not equally distributed in Vietnam. Areas where it was heavily used were generally combat areas that differed in terrain and flora from areas where it was little used. These areas may also have differed in other important respects, such as indigenous diseases, level of combat intensity, and type of personnel deployed. It is for these reasons that CDC proposes choosing the "likely exposed" and "likely not exposed" cohorts from the same area of Vietnam. Unfortunately, because of the inherent limitations of the records,, this approach may have the effect of increasing exposure misclassification (especially the categorization of those who were truly exposed into the "likely not exposed" group). These two competing forces, the desires for comparability and for maximum exposure separation, have drawn CDC to recommend a three-cohort design. Two of the three cohorts will be from the same area of Vietnam, III Corps (and the same time during the war, 1967-1968), but will differ in regard to their exposure likelihood. These two cohorts will be comparable but jnay suffer from imprecision of exposure separation. The third cohort will be drawn from other areas of Vietnam (but also from the same time period), areas where there is good evidence of little or no herbicide usage. This cohort will give maximum exposure separation from the "likely exposed" cohort but may suffer from a lack of comparability with respect to other health-influencing factors. This design is illustrated in the following 2 x 2 table which cross-classifies exposure by a measure of service experience. Likely Herbicide Exposure Yes No A Cohort 1 Cohort 2 Service Experience B Cohort 3 The empty cell, representing the combination of herbicide exposure with "Service Experience B," cannot be filled, because it is our understanding from the military that herbicide use was inextricably entwined with a certain service experience, as explained earlier in this paragraph. Because of the empty cell in the table, this design will present problems in analysis and interpretation. Moreover, the comparison of the first and third cohorts, which will ensure maximum exposure separation, may be subject to respondent bias; respondent bias should not be a problem in a comparison of cohorts 1 and 2, because individual respondents will probably be uncertain about their (study) exposure status. Despite these problems, we believe that this design is better than either of the other alternatives — alternatives based on an approach that uses only two cohorts—either decreasing exposure misclassification by decreasing comparability or increasing exposure misclassification by increasing comparability. The results of the Ranch Hand study, soon to be released by the U.S. Air Force, may help in the interpretation of this design. The Ranch Hand study will compare the health of crew members who flew the herbicide spray missions with air crew members who did not fly spray missions. Thus, it will provide information about herbicide exposure in the absence of the general experience of ground troops. �Page 8 3.2. Vietnam Exper ience Study The idea of studying health effects which might derive from the "general experience" of having been in Vietnam is at once attractive and unappealing. In part, CDC recommends this study as a "backup" for the Agent Orange study — if the Agent Orange study does not reveal any adverse health effects, veterans still will want to know if some other factors in their Vietnam service contributed to their perceived poor health. The major reason for CDC's recommendation is that there may have been many factors in addition to herbicide exposure which could have adversely affected those who served in Vietnam, in contrast to their counterparts who served elsewhere. It is also plausible that Vietnam veterans who did not see active combat in Vietnam were subjected to health-influencing events that were not part of the experience of those who served elsewhere. Any study which focuses on Agent Orange alone will obviously not test such a plausible multifactorial hypothesis. However, the multifactorial nature of this hypothesis makes the study of the "Vietnam experience" unappealing from the scientific point of view. The "experience" comprises numerous factors, many of which are unknown, poorly defined, or not quantifiable. Nevertheless, in our opinion, this is an important question to the Vietnam veteran and one that deserves as much attention as the issue of the possible effects of herbicide. Viewed in the broadest terms, the Vietnam "experience" could have influenced anyone who served there. A major concern about the validity of making a comparison of Vietnam and non-Vietnam veterans derives from an undocumented suspicion that there may have been preexisting differences between the two groups in terms of health-influencing factors and behaviors. If such differences existed and if they applied to all veterans, then a valid study of the Vietnam "experience" would not be possible. However, military personnel with whom we have consulted do not believe that such factors would have existed for all Vietnam veterans. Specifically, they believe that being sent to Vietnam was a matter of the "luck of the draw" (conditional on occupational specialty) for those who were in the Army and who were drafted or who were short-term enlistees. Serving in Vietnam, the U.S., in Europe, or elsewhere depended, in their opinion, on occupational specialty and the operational needs of the various commands. Thus, any given serviceman was at risk of serving anywhere there was a need for his occupational specialty. Individuals for the two cohorts of this study will be chosen on the basis of a review of randomly chosen personnel records located at the St. Louis records center. 3.3. Selected Cancers Case-Control Study As noted in Appendix B, several Swedish case-control studies (Hardell and Sandstrom, 1979; Eriksson et al., 1981; Hardell et al., 1981) suggest that soft tissue sarcomas and lymphomas occur 5-6 times more frequently in workers occupationally exposed to TCDD-contaminated phenoxyherbicides than in those not exposed. In addition, a National Institute for Occupational Safety and Health review of four U.S. company studies seems to have demonstrated an excess of deaths from soft tissue sarcoma among workers employed in plants where chlorinated phenols and their derivatives were manufactured (Honchar and Halperin, 1981). These studies have generated a specific concern among Vietnam veterans that they may be at increased risk for sarcoma and lymphoma, but no published studies address this question. CDC's proposed case-control study will determine if men who served in Vietnam are at increased risk of developing these tumors. In response to suggestions received from reviewers �Page 9 of CDC1s draft protocol, individuals with primary liver cancer and nasal and nasopharyngeal cancers will also be included in the case group. These are quite rare cancers and, in the absence of the hypotheses regarding sarcomas and lymphomas, would probably not deserve special study. Thus, the major focus of this study will remain on sarcomas and lymphomas. "Cases" will be males in the age range of Vietnam veterans identified by population-based cancer registries as having the specified tumors. Because of the study design, other cancers could be easily added if an association with phenoxyherbiclde exposure is suggested or if other evidence gives rise to specific concerns among Vietnam veterans. In this cancers will contribution in males (in large. study, information about other suspect risk factors for these be gathered. Thus, this study will permit an evaluation of their to the occurrence of these cancers, both in Vietnam veterans and the same age range as Vietnam veterans) in the population at �Page 10 4. Study Procedures 4.1. Selection of Study Subjects The selection of study subjects for the two cohort studies will be based on a review of military records by the Army Agent Orange Task Force (AAOTF) according to criteria set forth below. Selection of subjects for the Agent Orange study will depend on a simultaneous consideration of the position of U.S. troops in Vietnam and the times and locations of herbicide applications as indicated by extant records; neither the location of troops nor their proximity to herbicide applications will play a part in the choice of subjects for the Vietnam Experience study. Choice of subjects for the Selected Cancers study will derive from work done by CDC and the cancer registries participating in the study. Since this study is a case-control study, beginning with persons with the cancers and those without, military records will not be used as a part of the selection process, although they will be used as an aid to assessing exposure to herbicide among subjects who turn out to be Vietnam veterans. CDC intends to limit individuals included in all three proposed studies to men. The exclusive attention to males does not derive from a lack of concern about the health of those relatively few females who did serve in Vietnam. Rather, this decision is based on CDC's belief that if females are to be studied, they should be studied separately in sufficient numbers to allow meaningful conclusions to be reached about them as a group. Moreover, any study of women would require somewhat different sampling strategies and different emphases in interviews and medical examinations. CDC is concerned that a study of female veterans might be difficult to implement because of the probability that female veterans, once identified from military records, will be harder to locate than men because of the name changes which will have occurred because of marriages after discharge. The AAOTF and CDC are assessing the locatability of female Vietnam veterans. If this assessment proves that it is indeed possible to locate a sufficient proportion of them, CDC will design a separate study and prepare a protocol for review and possible funding. Such a study would probably most resemble the Vietnam Experience study proposed here for males, but a study of cancers similar to that proposed for males will not be possible — too few women served in Vietnam for any meaningful case-control study to be done. 4.1.1. Agent Orange Study • CDC proposes to limit this study to draftees and single-term enlistees in the non-officer ranks who served in the Army (grades El through E5 only); selection will be further limited to those who had only one tour of duty in Vietnam. Exclusion of officers is based primarily on a desire to make the groups as homogeneous as possible with respect to pre-existing demographic factors which could influence health. In addition, the inclusion of officers might require substantially increased record review to assess herbicide exposure potential (see below) because of multiple tours of duty in Vietnam. Exclusive focus on veterans of the Army is chosen for several reasons. The Army had a much greater proportion of draftees than the other services, and we believe that it is important to include substantial numbers of them in the study. Use of draftees will probably make achieving a balance on such �Page 11 factors as training, military occupational specialties, and pre-existing demographic factors easier. Inclusion of substantial numbers of draftees is also motivated by a desire to try to assess the possible association between volunteerism and health, (if, however, a large percentage of enlistees joined the Army because they felt that the draft was inevitable, such an assessment may not be possible.) CDC proposes to exclude the Marine Corps in part because its men were mostly volunteers and in part to limit the amount of records review required to select study subjects (the reasons for this will be better appreciated after the selection process is described). In addition, the AAOTF has worked most extensively with the records of the U.S. Army, has become most familiar with them, and is most confident about their quality. Moreover, the Air Force did not keep records that show the daily geographical placement of personnel, and rather limited numbers of Navy servicemen were stationed on land in the Vietnam theater. Even though all study participants will be males in the non-officer ranks who were in the Army, the results will probably be useful in making inferences about all men who had similar ground experiences and possible herbicide exposures in Vietnam; if there are no sex-specific effects, the same may be said about females. As noted previously, three cohorts of men will be chosen for the Agent Orange study. The first two, which will differ with respect to the likelihood of exposure to herbicides, will be chosen from III Corps (an area where herbicides were used extensively) during the same period of time, 1967-1968. This will be done to make the two as similar as possible with regard to the nature of their service experience — similar with regard to, for example, type of terrain, indigenous diseases, and intensity of combat. To enhance the possibility of including soldiers who may have been exposed to herbicides, we will select the men included in these first two cohorts exclusively from combat battalions. Since these two cohorts will be chosen from an area where herbicides were extensively used, there is a potential for exposure misclassification. The third cohort will therefore be chosen from an area where there is good evidence that herbicides were not used. According to the AAOTF staff, this third cohort probably cannot be exclusively derived from combat battalions. Veterans to be included in the first two Agent Orange study cohorts will be selected by a multi-step review of military records, beginning with the selection of a geographical area of consideration and ending with the choice of individual soldiers. Since many of the proposed procedures are untested, modification, indeed even a recommendation not to proceed with an Agent Orange study, may be required after pilot study assessments (see section 4.5.1.1. below). In summary, the steps required are: 1) 2) 3) 4) 5) select a geographical area and time of interest - this will be III Corps and 1967-1968 determine which of the battalions stationed in III Corps in 1967-1968 have acceptable records choose a random sample of 50 battalions (250 companies) from among all battalions with acceptable records choose 2 random subsamples of 25 battalions (125 companies) each from the 50 battalions chosen in step 3 abstract selected companies' locations for subsample 1 on all days in 1967-1968 �Page 12 using the "Herbs" and "Services Herbs" tapes, score the herbicide encounters of the 125 companies of subsample 1 on all days 7) rank the 125 companies of subsample 1 with respect to their cumulative herbicide encounters 8) choose men for the "likely exposed" cohort from companies at the top of the ranked list and men for the "likely not exposed" cohort from those at the bottom of the list 9) using the "Herbs" and "Services Herbs" records, score individual men chosen in step 8 for their herbicide encounters according to the scoring schemes used for battalions (step 6). 10) repeat steps 5 through 9 for the 25 battalions comprising subsample 2, with the following modification: rank herbicide encounters using the 125 companies of the 25 battalions of subsample 2 and the companies of subsample 1 which were not chosen in the first iteration of step 8 6) The rationale for these steps is presented below. To limit the amount of records review required, we first restricted, on the advice of the AAOTF, the geographical area of consideration to 111 Corps and the time period to 1967-1968. This area and time period were selected because of a variety of factors, including the number of Ranch Hand missions and U.S. troop strength, which was near peak. The AAOTF has determined that about 110-120 Army combat battalions were stationed in III Corps at some point during that time (usual battalion strength was 1,000). The records of the companies attached to battalions determined to have served in III Corps will be the major source of information about troop locations. The second step in the selection process will consist of a review of General Services Administration (GSA) documents to ascertain which battalion records appear to have unacceptable time gaps (if gaps appear in battalion records, it may be possible to supplement them with division and brigade level records, and this will be done when feasible). CDC does not believe that it is necessarily wise to exclude a unit simply because some of its records are missing — units with missing records could have had more or less exposure to herbicides than units with complete records. Therefore, CDC proposes to apply the following criteria regarding records quality: if a battalion has more than 30 contiguous days of absent records or an aggregate of more than 60 days1 absent records for the period 1967-1968, the unit will be considered unsuitable for inclusion in the study. If very few units are found to have gaps of this magnitude, more stringent criteria can perhaps be used. For each of the combat battalions located in III Corps in 1967-1968, the AAOTF will summarize the condition of the records as indicated in the GSA documents. The third step will be the choice of a random sample of 50 battalions (250 companies) from among those judged suitable during the second step. Step four will involve splitting the sample of 50 battalions into random halves of 25 battalions each. Fifty battalions will be sampled in order to limit the quantity of records review required, but this sampling should provide a reasonable estimation of the range of herbicide encounters (next paragraph). CDC believes that this is an important issue — at this point the frequency and nature of troop herbicide encounters is largely a matter of conjecture. As noted before, the records available will never permit an unambiguous assessment of exposures, but this approach will help to place a frame of �Page 13 objectivity around the issue, at least for men in Army combat units in III Corps in 1967-1968. Step five will involve abstracting from company records (or battalion records, if necessary) all locations recorded for the selected companies on each day during 1967-1968. The purpose of dividing the sample of 50 battalions into halves is to increase the speed with which CDC can proceed with the interviews and examinations. The AAOTF estimates that it will take 18 months to abstract location information for all 50 battalions and that it will take 12 months to do it for the first 25 battalions. Step five will involve abstracting from company records (or battalion records, if necessary) all locations recorded for the selected companies on each day during 1967-1968. In step six, CDC will check the company locations against the locations of herbicide applications as recorded on the "Herbs" and "Services Herbs" tapes. The "Herbs" tape contains computerized records of Ranch Hand missions (time, place, type, and amount of herbicide). The National Academy of Sciences report (1974) on the effects of herbicide usage in Vietnam contains a relatively limited assessment of the accuracy of these records. CDC finds the results of this investigation encouraging, but doubt about accuracy exists in some quarters today. CDC has requested that the National Academy make available the results of other checks done at the time and that it look into the possibility of further accuracy checks. The "Services Herbs" tape primarily contains records of non-Ranch Hand herbicide applications (e.g., base perimeter sprayings). This set of data has been put together by the AAOTF from a review of a variety of military records; the degree of completeness of the "Services Herbs" data set is unknown. The number of unit encounters with herbicide applications according to these data sets will be tabulated by at least three systems; other systems may be used if this seems warranted. The first of these systems will have geometrically progressing scores or weights for various space and time distances, and the second will have linear weights. The aggregate scores for these two systems will be based on the products of the time and space scores. The third system, a variant of one proposed by the Department of Defense, will simply count the number of encounters which are at distances of less than 3 days and 2 kilometers. The purpose of these exposure systems is to obtain a spread of unit exposures so that units can be chosen from the top and bottom of the scales. It is desired that the spreads obtained should reflect "meaningful" differences in exposure. Relatively little is known about the environmental fate of herbicides and TCDD, and even less is known about the human pharmacokinetics of these substances. Because of this lack of knowledge, these systems are necessarily arbitrary and this motivates the proposal of three scales. The scorings for the first two systems proposed for preliminary tabulation are indicated below. Exposure System A. 1. Ranch Hand Missions a. Regular Missions — cross-classified by time after mission «»1 day, score=16; 2-3 days, score=4; 4-30 days, score=2; and 31-59 days, score»l), distance (<-l km, score-4; 2-3 km, scored; 4-8 km, score-1), and type of herbicide. �b. 2. Page 14 Aborted Missions — cross-classified and scored as above. Other Herbicide Applications (e.g., perimeter spraying)—for those encounters O 1 km classified by time and scored as above Exposure System B. 1. Ranch Hand Missions a. Regular Missions — cross-classified by time after mission «»1 day, score=4; 2-3 days, score«=3; 4-30 days, score»2; and 31-59 days, score-1), distance (<-l km, score-3; 2-3 km, score«2; 4-8 km, score-1), and type of herbicide. b. Aborted Missions — cross-classified and scored as above. 2. Other Herbicide Applications (e.g., perimeter spraying) — for those encounters <- 1 km classified by time and scored as above. As mentioned before, the various encounters will be weighted by the product of the time and distance scores; each encounter of a unit with a particular herbicide application will be counted in only one time and one distance category. For example, using Exposure System A, an encounter with a Ranch Hand mission within 1 day and 1 km would receive a score of 64, as would an encounter with a base perimeter application within 1 day (small bases); an encounter with a Ranch Hand application within 4-30 days and 2-3 kilometers would get a score of 4. Using the third (modified Department of Defense) system, any encounter which occurs within the 3 day-2 kilometer limit would receive a score of 1. For each of the 3 exposure systems, the daily scores will then be summed over all days in 1967-1968 for each company. Next, the 125 or so companies of subsample 1 will be ranked on their summed encounter scores. If there is good agreement in the rankings provided by the three systems, those at the top of the lists will provide individuals for the "likely exposed" cohort and those at the bottom will contribute to the "likely not exposed" group. If there are substantial disparities in the rankings provided by the three systems, then roughly 1/3 of each of the two cohorts will be chosen from the top and bottom of each of the rankings. At this time it is unclear how many companies will have to be selected to provide the requisite number of individuals for these 2 cohorts, but it will probably be on the order of 50 to 60 from the top and a like number from the bottom (of both subsamples combined). If 55 companies each provide 150 suitable Individuals, this number will allow some loss because of non-participation and will yield the number desired for each of the cohorts (see section 4 4 1 ) .... About 40% of the final sample of men will be derived from subsample 1, and the remainder will be chosen after the re-ranking in the second iteration of steps 5 through 8. The desire to omit the Marine Corps from this study can now be more easily understood. If Marines were included, the records review and other selection tasks to this point would have to be done separately for them because they were largely stationed in I Corps, and this would cause delay. The next step will be the choice of individual soldiers from the selected units. This process will begin with a review of company morning reports. �Page 15 Individuals who appear to meet the criteria with respect to type of entry into the service (draftee or single-term enlistee), are in the non-officer ranks, and whose 1- year Vietnam tour began and ended during 1967-1968 will be considered potentially eligible for inclusion in one of the cohorts. For those who appear to be eligible, the AAOTF will also document their presence or absence with the selected units on each of the days during the 2-year period 1967-1968. Those individuals who were absent from their units for more than 90 days of their scheduled 12-month tours (exclusive of their regular R&R leave) will be considered ineligible for final selection. The AAOTF will also document the reasons for all absences for both the selected men and those men who would be eligible except for their absences. Thus, this process will provide CDC with, inter alia, a measure of combat intensity, since absences for reason of casualty will be recorded. Individual personnel folders will be obtained by the AAOTF from the National Personnel Records Center in St. Louis for soldiers considered eligible. The AAOTF staff will abstract certain identifying and service (e.g., military occupational specialty) information from the individual personnel folders and forward the information to CDC on an incremental basis so that CDC can begin the process of locating the veterans and soliciting their participation in the studies. Later, in step 9, individual soldiers will be classified with respect to exposure to herbicides by a scheme similar to that noted above. At this step some men drawn from units at the upper end of the exposure scale may be found to have "low" individual exposure scores and vice versa. Such men might be assigned to the other cohort or they might be omitted from the study altogether. The third cohort for the Agent Orange study will be selected by a different method. Areas in Vietnam where there would have been no reason for herbicide usage will be identified by the AAOTF and a roster of units which served in, and only in, those areas in 1967-1968 compiled. The staff of the AAOTF has suggested that Cam Ranh Bay or Vung Tau might be examples of such areas. CDC will check the locations of these areas against the herbicide usage records to ensure that there was no herbicide use. Enough units will be randomly chosen from this roster for the required number of individuals to be included in the study. The eligibility criteria for selecting individuals from within the selected units will be the same as those used for the first two cohorts. The AAOTF will provide CDC with the same sort of identifying, service, and absence information that it provides for those individuals included in the two other cohorts. 4.1.2.Vietnam Experience Study The procedures for selecting individuals for the Vietnam Experience study will be substantially different from those used for the Agent Orange study — the process will start with the selection of individual personnel files in the National Personnel Records Center in St. Louis rather than with the selection of military units. We understand that, for draftees and single-term enlistees in Army combat units, assignment to Vietnam or to some other part of the world was essentially a random process, but this was probably not the case for other services. Since the desire is to compare men who went to Vietnam with men who did not, but who had a more or less equal chance of being assigned to Vietnam, CDC also proposes to limit this study to Army veterans in the non-officer ranks (grades El through E5). �Page 16 The St. Louis records center houses personnel files for all discharged service persons, except the living retired and those in the active reserves. Soon after discharge, the military personnel folder is transmitted to the center where it is identified by service and given an accession number. Since a master list by service and accession number is available, a sample of individuals can be selected from the records center stacks. Unfortunately, the master accession list does not indicate whether the discharged soldier served in Vietnam or not, his rank, or any other vital information. Thus, the records of each individual identified from the accession list will have to be pulled to determine if he qualifies for inclusion in the study. This eligibility assessment will be done at the records center and coordinated by the AAOTF staff; records of individuals found to be eligible at this preliminary review will be sent to AAOTF headquarters in Washington, D.C., for complete review. CDC staff have visited the St. Louis center and reviewed a systematic sample of 101 Army personnel records. The records were chosen to encompass those accessed by the Center from 1966 through 1973. Of the 101 selected, 1 was missing, 3 were checked out, and the contents of 4 could not be interpreted by CDC staff. Sixty-one of the remaining 93 were single-term draftees and enlistees; 24 of the 61 single-term soldiers served in Vietnam, 10 served in Europe, 8 in Korea, 16 in the U.S. only, and 3 elsewhere. This work indicates that the approach can yield a sample with relatively little wasted effort, and CDC believes that it is far preferable to a sampling scheme based on a preliminary selection of military units. The members of both cohorts for the Vietnam Experience study will be selected from among those soldiers whose personnel folders were acquired by the records center during 1965-1977; those chosen will have entered military service in 1965-1971 and will have served in Vietnam during the years 1966-1972. For the Vietnam service cohort this should provide a year-of-tour distribution roughly proportional to the year-by-year Army troop strength in Vietnam over the period 1966-1972. The selection procedure for the control cohort will be such that its period of service distribution is equivalent to that of the Vietnam cohort. The cohort of men included in the Vietnam service cohort will have served only in the U.S. and Vietnam. It is proposed that the control or non-Vietnam cohort be chosen so that it comprises three groups: (1) men who served only in the continental U.S.A., (2) men who served in the U.S.A. and Europe, and (3) men who served in the U.S.A. and Korea. The numbers of men in these three groups will be proportional to the military strengths in the three areas in 1966-1972. AAOTF will give CDC the same sort of information about each soldier in this study as will be provided for those men in the Agent Orange study, except that no daily geographical location information will be given. 4.1.3. Selected Cancers Case-Control Study As noted before, this part of CDC's efforts to address concerns of Vietnam veterans will take the form of a population-based case-control study. A case-control study is recommended because a cohort study would require truly massive sample sizes to detect an increased risk for such rare diseases — much larger samples than those proposed for the Agent Orange and Vietnam Experience studies. Studying such large samples would unnecessarily delay CDC's ability to provide answers to veterans about their risks for more common disorders. �Page 17 The term population-based implies that all cases of the selected cancers in defined population groups will be ascertained and an attempt made to include them in the study. This will confer at least two major advantages over studies done with cases collected by other methods: 1) since all cases arising in a population are ascertained, the concerns about biases of ascertainment which always attend other case-selection strategies are not at issue, and 2) a population-based study allows estimates of attributable risk, not just relative risk. The control group will be chosen from the same population as the case group, and this will allow disease incidence rates to be estimated by veteran status. It is proposed to use the Surveillance, Epidemiology, and End Results (SEER) Centers, sponsored by the National Cancer Institute, as the major source of cases. The SEER Centers ascertain nearly all people newly diagnosed with cancer in 10 defined population areas (National Cancer Institute, 1981). These areas are: the states of Connecticut, Hawaii, Iowa, New Mexico, Utah, and the Commonwealth of Puerto Rico; and the metropolitan areas of Atlanta, Detroit, San Francisco, and Seattle. CDC has contacted eight of the SEER Centers by telephone and they have indicated that they are interested in participating. Overall, interest in participation appears high because the SEER centers want to continue to build and demonstrate their epidemiologic potential. In addition, each center employs at least one epidemiologist, many of whom have been involved with the issue of cancer and chemical exposures and who view the proposed study as personally interesting. Overall, CDC believes that the SEER network is a superb epidemiologic resource that has been proven in other large case-control studies, such as those that investigated the association of bladder cancer with artificial sweetener use (Hoover and Strasser, 1981) and uterine, ovarian, and breast cancer with oral contraceptive use (Layde et al., 1983). Other population-based cancer registries may be used for case ascertainment, if they are interested in collaborating in this study and if their case ascertainment is complete and rapid enough. All cases of the selected cancers occurring from July 1, 1984, to June 30, 1988, in males with birthdates 1929-1953 who reside in the geographic areas covered by the participating population-based cancer registries will be included in this study; the "cases" will be contacted and interviewed within 6 months of diagnosis. Men in this age group have been selected because they were of military service age during the years herbicides were used in Vietnam (see section 4 4 2 . Since soft tissue sarcomas are so rare, CDC has ..) considered including additional cases diagnosed before July 1, 1984, in order to increase the power of the study to detect any association which may be present between herbicides and/or service in Vietnam and sarcomas. This possibility has been rejected for three reasons. 1) Most importantly, the Swedish studies which suggest a relationship between sarcomas and occupational exposure to 2,4,5-T indicate a mean latency period between first exposure and diagnosis of about 16 years. Therefore, including cases that arose before 1984 might give only an illusion of increased power. 2) Because the fatality rate for soft tissue sarcoma is quite high (Tucker and Fraumeni, 1982), information about early cases and controls would frequently have to be gathered from next-of-kin instead of from the affected man. However, this latter point would not be a major concern if data collection for these cases were limited to relatively simple items, such as whether the man served in Vietnam. 3) The New York State Health Department has completed a sarcoma study for cases diagnosed in 1962-1980 and the VA and the Armed Forces Institute of Pathology are planning a study directed at cases diagnosed in 1975-1980. �Page 18 Four histologic review panels, each composed of 2-3 pathologists, will be established—one group to review each type of cancer. The groups will receive a set of slides or tissue blocks on each case and will establish their own diagnosis without knowledge of the presumed diagnosis. Interviews with cancer cases will not be delayed for confirmation by the pathology review panels. Controls will be selected by the method of random digit dialing (ROD). Telephone numbers are randomly phoned, and a brief census of the household is made. If a man of the right age is found, then he will be asked to participate in the study. This method worked successfully in the National Cancer Institute (NCI) Bladder Cancer Study (Hoover and Strasser, 1981) and CDC's Cancer and Steroid Hormone Studies (Layde et al., 1983). Over 90% of the households that had eligible women in CDC's study yielded an interview; the NCI results were similar. Unlike the usual methods of collecting a sample of a population, which depend on making at least a partial in-person census of the geographic area, ROD allows this to be done by telephone, which clearly is less expensive and far more practical. About 95% of households have telephones. In addition, as detailed in Appendix C, several researchers have documented how well samples chosen by RDD reflect the general population. The main concern is that people of very low socio-economic status may be underrepresented in the control group. CDC believes that the effect of this potential bias will be small for two reasons: 1) our control group will be so large that some very poor people will be included; and 2) an analysis stratified by socio-economic status should help ameliorate whatever bias is present. On the basis of the age and race distribution of cases, CDC will select controls from the list of eligible men so that the overall age and race distribution of the controls will be similar to that of the cases. As the study progresses, if the age distribution of cases is different from what is expected, control selection can be modified. 4.2. Location of Study Subjects For each of the veterans selected for the Agent Orange and Vietnam Experience studies, CDC will receive from the AAOTF a variety of identifying information with which to begin the location process. The information available for each man will, in addition to his full name, include: his Social Security Number (SSN) and service number; the address he gave the military at discharge; the name and address of one parent and the name and address of one sibling (the names and addresses of relatives are not invariably available in the records). Although this may seem to be a substantial amount of information with which to begin tracing, the addresses will be about 15 years old, and CDC expects to experience great difficulty in locating individuals — indeed CDC believes that this could present such a formidable obstacle that it may not be possible to complete these studies using the sample selection strategies proposed here (see section 4.4.1. regarding minimum acceptable participation rates and section 4.5.1. for a discussion of the role of the pilot studies). If it should turn out that these two cohort studies are not feasible, CDC would propose another plan for the Vietnam Experience study, but an Agent Orange study should start with military unit records. The alternative plan for the Vietnam Experience study would involve sample selection by a variant of the RDD technique described in section 4.1.3 for the Selected Cancers Study. However, this alternative plan would involve considerable expense in identifying the requisite number of veterans. �Page 19 The Air Force's Ranch Hand Study team had great success in locating its study subjects — 97% for the Ranch Hand group and 93% for the control group. This gives CDC a standard to reach for, but there will be marked differences between the Ranch Hand subjects and the subjects selected for CDC's two cohort studies. About 25% of the Ranch Hand sample was still on active military duty at the time data were collected and another 25% was composed of men retired from the Air Force (and therefore receiving pension payments). Thus, the location of about 50% of the Ranch Hand study sample was known before the study began. Very few of the men selected for the Agent Orange study are expected to be on active duty at this time, and none of the Vietnam Experience study subjects will be, because they are to be chosen from the St. Louis records center (section 4.1.2). The one reason for optimism is that SSNs will be available for virtually all those chosen for the two cohort studies. CDC expects that the major locating source will be the Internal Revenue Service (IRS). CDC will submit the names and SSNs of the desired veterans to the IRS which will return to CDC the most current addresses available. This should be a very good source, but there are inherent limitations. Most importantly, the IRS has current addresses only for persons who have recently filed tax returns; IRS will remit addresses for individuals who have not paid taxes for some time, but it will not indicate whether the addresses are current. It is obvious that if some veterans (or more importantly the aggregate of veterans in one of the study groups) are operating on the margin of economic life, they will be difficult to locate. The SSNs will also be transmitted to the Social Security Administration (SSA), which can let CDC know if a man is deceased and, if not, if he has recently been paying social security taxes and who his employer has been (CDC experience in using SSA records for tracing indicates that the records used for this work may be out of date by 2 or 3 years). SSNs may also be given to the Veterans Administration which can check to see if a death benefit has been paid. Furthermore, the SSNs will be used for future mortality followup (see section 4.3.1.1) through the National Center for Health Statistics* (NCHS's) National Death Index. If the simple approaches described above fail to locate a study subject, then much more labor-intensive, difficult, and expensive procedures must be used. These procedures will almost certainly involve field "detective" work and the use of such sources as credit bureaus and contacts with neighbors at the last address of record. Because of the design of the Selected Cancers Study, CDC does not anticipate that the location of study subjects will present significant problems. 4.3. Ascertainment of Health and Exposure Status A variety of health and exposure data will be collected for each of the participants in the two cohort studies and in the Selected Cancers Study. The categories of items to be collected and the methods by which they are to be gathered are presented below; Appendices D-E contain relatively specific topical lists of items of interest. The specific items to be included in questionnaires and examinations may be modified because of new findings from studies now in progress (e.g., Ranch Hand; see also section 4.6.1). �Page 20 4.3.1. Agent Orange and^Vietnam Experience Studies 4.3.1.1. Mortality In forma tion It is projected that the first component of both cohort studies to be completed will be mortality assessments. It is proposed that mortality assessment of the five different cohorts be repeated every 5 years for an indefinite period of time through use of NCHS's National Death Index. During the main studies, the fact of death will be ascertained in the course of attempts to locate the selected veterans (section 4.2). As noted before, the name and SSN of any study subject who does not appear on the returns from the IRS or who cannot be located will be submitted to NCHS, SSA, and VA. The NCHS can provide help through the National Death Index for those who died after 1980. SSA or VA should also be able to indicate which veterans are deceased and, in addition, may be able to provide locating leads for subjects who are still living. The VA's Beneficiaries Identification and Records Location System (BIRLS) files will be particularly useful in identifying veterans who died before 1981. In 1981, veterans' burial expenses provided by the VA were reduced, and there may have been reduced reporting of veterans' deaths to the VA after that change. In addition, some deceased may be identified by relatives or neighbors who are contacted during the location process. During the study CDC will estimate the degree of underascertainment of deaths by extensions of the capture-recapture methods used by ecologists (Hook and Regal, 1982). There are unlikely to be enough deaths among veterans in the pilot sample, however, to assess accurately the completeness of identification of the deceased before the full-scale study. Once the fact of a death has been ascertained, CDC will proceed to obtain records that will help to establish the cause. Death certificates will be routinely obtained, usually from the vital records department of the state in which the death occurred. In order to provide the most powerful assessment of mortality, it is important to have accurate accounts of the causes of death, and death certificates suffer in this regard — they are only accurate for rather broad cause groupings. This quest for accurate cause-of-death information is considered to be particularly important at this point, since the numbers of deaths in this group of men is, on the basis of U.S. mortality statistics, expected to be small (Table 1). Therefore, when possible, hospital records, autopsy reports, and other documents that will help establish the cause of death will be obtained. Mortality data will be analyzed in two ways: first using only death certificate information, and second using the supplemented certificate data. In the course of selecting the cohort members, those who were killed in action will be ascertained; this will give us one measure of the combat intensity to which members of the various cohorts were subjected. 4.3.1.2. Morbidity and Exposure Information Data regarding the morbidity experience of the study subjects will be collected through health interviews and medical and psychological examinations and through selected laboratory tests. �Page 21 4.3.I.2.A. Health Interviews CDC proposes to conduct personal interviews with all study subjects who agree to participate in the studies (6,000 per cohort). These interviews will be conducted by telephone by specially trained interviewers. Telephone interviews may be supplemented by in-person interviews if the pilot study indicates that participation may be suffering because too few study subjects can be reached by phone (about 95% of U.S. households have phones). It is anticipated that the interviews will be done by using a "computer assisted telephone interviewing" (CATI) system. CAT1 has numerous advantages over the traditional paper and pencil telephone or in-person system. Most importantly, CDC believes that much better quality control is possible when a CAT1 rather than a traditional system is used. Examples of the enhanced quality control include data checking and editing while the interview is in process, modification of the questionnaire to fit the individual respondent, automated implementation of interview skip patterns, and the ability to monitor the interviewers' transcription of respondents' answers to questionnaire code ( . . the interviewers' video displays can be watched on a monitor, by an ie, authorized supervisor, at the same time audio monitoring is done). It is hoped that two interviewers can be used to conduct each veteran's interview. One interviewer will ask questions about military service and other exposure-related matters and the second will ask questions about health and other outcome-related issues. The purpose of using two interviewers is to keep the interviewer questioning about health "blind" to the "exposure" status of the veteran being interviewed. During the next few months, CDC staff and outside consultants will design the formal interview instruments, including the detailed wording of questions. The general types of questions are explained below, and a topical list of items to be Included in the Interviews can be found in Appendix D. Questions will be asked about a wide variety of health outcomes and also about exposures and behaviors which may predispose to ill health. Some variables in the latter category may be confounding factors — factors which may be associated both with health outcomes and with exposure (cohort) status. For example, race is a risk factor for many diseases and may be associated with cohort membership. If the proportions of blacks and whites in the several cohorts are not equivalent, a race effect could be confounded with, or mistaken for, a cohort effect for any health outcome where race is a predisposing factor. Therefore, race needs to be ascertained during the interviews so that if an imbalance is present, it can be accounted for during data analysis (section 4 6 2 . In addition, a limited number of questions ..) will be asked about each subject's military experiences. Apart from basic administrative data, we have categorized the items to be included in the interviews into four categories. Examples from each of these groups are presented below, along with a brief rationale for collecting such information. —Sociodemographic Information Variables in this class include race, place of residence, marital history, problems in obtaining employment, occupation, income, and education. Most of these variables are potential confounding factors, as discussed above, and are therefore required for analysis. In addition, some of these social �Page 22 characteristics are themselves possible effects of service in Vietnam and are therefore of interest as psycho-social outcomes. —Medical History This area forms the heart of the interview. The concerns veterans have expressed about their health have been wide ranging — numerous types of complaints have been heard. There are no strong hypotheses which can guide our inquiry, and it must be therefore thought of as being essentially descriptive. However, there are certain pointers from animal experimentation, from industrial exposures, and from the lay press which can guide us so that we do not overlook areas of concern. And our regular monitoring of comparisons between the various cohorts for major health outcomes will allow us to generate specific hypotheses and supplement or expand on certain lines of questioning as the study progresses (see section 4 6 1 . In addition to ..) the standard closed-ended questions about major health outcomes, the interview will provide an opportunity for open-ended responses to queries about what concerns individual respondents have about their health. These answers will be monitored at regular intervals so that anything striking can be included in interviews with later respondents. In the Agent Orange study more emphasis will be given to dermatologic and immunologic outcomes, whereas in the Vietnam Experience study more emphasis will be given to psychologic outcomes. —Environmental and Occupational Exposure Information A wide variety of potentially harmful exposures are included in this class. Examples include those questions about occupational exposure, particularly to herbicides, smoking, alcohol, and illicit drug use. Some of these factors are accepted as risk predictors for certain diseases, but while some are only suspected. In addition, some of these factors may be associated with service in Vietnam, and therefore are potential confounders. —Military History A substantial amount of information about study subjects' military service will be available among the data provided to CDC by the AAOTF, but many important items will not. Specific areas which will require inquiry during the interview include an inquiry about occupational duties while in the military (to supplement the military occupational specialty designation which will be provided by the AAOTF), a scale to rate the intensity of combat to which individuals were exposed, and the study subjects' perceptions about exposure to herbicides. The combat scale will not be applicable to interviews done with the non-Vietnam service cohort included in the Vietnam Experience study, nor will questions about perceptions about exposure to fixed-wing herbicide applications. Two additional comments need to be made regarding the development of the questionnaire. First, because of the varied educational and cultural background of the veterans, care will need to be taken to ensure that participants understand all the questions. Second, the order of the inquiries on the questionnaire will not necessarily reflect that of Appendix D. Both the wording and structure of the questionnaire will be extensively evaluated during the pretest and pilot phases (see section 4 5 1 . ..) �Page 23 4.3.1.2.B Medical and Psychological Examinations; Laboratory Tests A random subset from each of the five study cohorts will be selected for participation in the medical, psychological and laboratory work-ups; the goal is to complete examinations on 2,000 men per cohort. The examinations will take about 2 days to complete and will be done in as few centers as feasible to minimize problems of standardization of methods among the centers. CDC would prefer one or two examining centers, but the availability of contractors capable of the necessary through-put is unknown; moreover, travel to distant locations may enhance or detract from obtaining a reasonable level of participation (see section 4 5 1 2 . The selection of subjects for each of ...) the centers (if there is more than one) could depend upon geography, or the selection could depend on which study the individuals are participating in, but this cannot be specified until the pilot studies and pretests are complete. It is hoped that one laboratory can be used to perform most tests. The items to be included in the examinations and the laboratory tests to be used are listed in Appendix E; as explained in section 4.6.1, this list could be modified, if indicated, by the results of the interviews or the early examinations. The lack of strong hypotheses mentioned above makes a relatively wide-ranging battery of tests and procedures necessary. In addition, the medical examinations and laboratory tests will be of high quality and fairly comprehensive as a service to the study subjects and to enhance the chance of achieving a high participation rate. Because of specific concerns about psychological disorders, especially post-traumatic stress disorder, a fairly extensive psychological and neuropsychological battery of tests will be used. The guiding principle in the choice of tests in this area was the need for well-standardized tests that yield numerical, not just qualitative, data. The neuropsychological tests measure visual and auditory perception deficits, learning and memory impairments, and attention, coordination, and dexterity abnormalities. The psychological tests focus on personality assessment, current symptomatology, and a standardized diagnostic screening procedure. To detect neurological and immunological deficits, some rather specialized procedures will be included. However, CDC'8 general approach will be to limit the examinations and tests to those that measure health and well-being deficits in the simplest and most direct way possible. For example, fertility problems will be evaluated in the interview (above) and in the history taken at the time of the examination rather than by the examination of sperm morphology and motility or gonadotropin assays. Only if the interview data suggest an average deficit in fertility in one or more of the cohorts will more elaborate testing be undertaken (section 4 6 1 . The ..) CDC study team also takes a skeptical attitude to such esoterica as the examination of peripheral blood cells for chromosome breaks — in this case one is at a loss to know what prognostic significance can be attached to chromosome breaks and other such abnormalities. If a test does not help a physician to make a diagnosis or if it does not itself indicate outcomes are associated with health and well-being or longevity, then the test will not be used. However, if more sensitive, specific, and reliable tests for the outcomes of interest become available during the course of the study, we will consider their feasibility and use in random samples of those selected for physical examination and laboratory testing. �Page 24 All medical examinations at each center will be done by physicians trained in appropriate specialties. When necessary, the examining physician will consult with another specialist (e.g., a neurologist or dermatologist). Examinations for which there may be substantial inter-observer variation will be done by one examiner at each center. The various examiners will be "blind" as to which cohort individuals belong. Quality control for laboratory tests will be done by the contractors' laboratories and monitored by the CDC staff. Study participants will be informed of the results of the examinations for those items where such knowledge will be of benefit to the individual veteran. (Some tests, particularly in the psychological area, may have little meaning for the individual because they are not designed for the purpose of making individual diagnoses.) If the study examinations raise suspicion about disease and extensive diagnostic work-up is required for definitive diagnosis, then the individual will be informed of the need and referred to the health-care provider of his choice, with copies of the pertinent portions of the evaluation. In such cases, CDC does not propose to complete definitive diagnostic workups, since this is more appropriately coordinated by the physician who will be caring for the veteran. 4.3.2. Selected Cancers Case-Control Study The information to be gathered in this case-control study is outlined below, and a detailed topical list is found in Appendix F. As for the two cohort studies, the actual interview instruments will be prepared over the next few months. CDC prefers that interviews for this study be done by telephone from a central location, using CATI (see above). If this is done, then the interviewer who collects most of the interview information can be "blind" as to the respondent's case/control status. However, participation by the various cancer registries will probably not be high unless they can use their own staff to do the interviews (this was the approach CDC used in its Cancer and Steroid Hormone Study). If the latter approach turns out to be necessary, then the use of CATI may not be feasible, although CDC will explore the possibility of implementing a CATI system on a microcomputer. Since survival is short for some of the cancers included in this study, in some instances next-of-kin may need to be interviewed. Information which will be gathered about known or suspect risk factors for the selected cancers is divided into five major groups. Examples from each of these groups are presented below, along with a brief rationale for collecting such information. In addition to the information about military service which will be collected during the interviews, the AAOTF will assist in making an estimate of the herbicide exposure likelihood for each Vietnam veteran case or control (AAOTF will not know the case/control status of the individual veterans when making this assessment). The exposure likelihood estimation process will be similar to, but much simpler than, that proposed for the Agent Orange study. The technique is similar in that it will depend on the proximity of individuals in time and space to herbicide applications. It is simpler in that the specificity with which this proximity is to be measured will be much lower than that proposed for the Agent Orange study. Specificity will be less because the records review needed to duplicate the Agent Orange technique would be especially burdensome — the veterans in this study could come from any one of the four branches of the military and from any unit stationed in �Page 25 Vietnam. The simplified technique is being developed by CDC and AAOTF for CDC1 8 birth defects study. — Sociodemographic Information The type of data and rationale is essentially the same as that for the Agent Orange and Vietnam Experience studies (see above). History of Cancer Soft tissue sarcomas and lymphomas have been reported to cluster in families. This tendency may be genetic or may reflect a persistence of adverse environmental circumstances in families, or both. The tendency of cancers to recur in families is not likely to be strongly related to service in Vietnam and therefore should not confound the analysis of cancer risk associated with that service. However, the risks of familial occurrence are not well known in the U.S.A., and this information will be useful for other reasons. History Underlying diseases which may predispose to the development of these tumors include rheumatoid arthritis, other cancers, celiac disease and gluten enteropathy, radiation or immunosuppressive therapy, diphenylhydantoin therapy for lymphomas (Grufferman, 1982; Greene, 1982), and immunosuppressive and radiation therapy for soft tissue sarcomas (Tucker and Fraumeni, 1982). Primary and acquired disorders of the immune system have frequently been associated with the development of these tumors. A medical history with specific questions regarding these risk factors will be included in the questionnaire. In some situations additional medical information may be needed to establish with certainty the underlying diagnosis. On an as-needed basis, the cancer registries will be responsible for retrieving additional information on the medical evaluation of these underlying medical disorders, including workup, histologic diagnoses, and/or histologic specimens. — Environmental and Occupational Exposure Information A wide variety of potentially harmful exposures are included in this class. Examples include those questions about occupational exposures, contact with animals, smoking, and illicit drugs. Some of these factors are accepted as risk predictors for cancer, but some are only suspected of being such. The following chemicals may be related to soft tissue sarcoma: arsenicals, vinyl chloride, and iron dextran injections (Tucker and Fraumeni, 1982). Halomethane, lead, asbestos, and cadmium may be related to lymphomas (Grufferman, 1982; Greene, 1982). In addition, some of these factors may be associated with service in Vietnam (e.g., alcohol or drug abuse hepatitis exposure) . — Military History Information collected about the military service of the cases and controls included in this study will be similar to that collected during the two cohort study interviews. �Page 26 4.4 Sample Sizes, Statistical Power, and Participation Rates 4 4 1 Agent Orange and Vietnam Experience Studies ... The sensitivity (power) of these studies to detect a real increased risk among the veterans in any one of the cohorts depends on several factors, most prominently the numbers in each of the cohorts, the prevalence or incidence of the condition of concern, the amount of misclassification on the variables used to define the cohorts, and the magnitude of the increased risk. It is proposed that each of the cohorts included in the mortality follow-up and health interview phases of these studies be composed of 6,000 men. The number 6,000 was chosen since this will give good power (beta^alpha=0.05, 1 tail) to detect a 2-fold increase in the risk for health outcomes normally occurring at the rate of about 5 per 1,000 in comparisons of two cohorts (if there is little or no misclassification in the selection of men for the cohorts) (see Table 2). A high beta level, equal to the alpha level, is suggested since CDC believes that as much attention should be given in these studies to type II errors as to type I errors. CDC further recommends that a sample of 2,000 be selected from each of the cohorts for the medical, psychological, and laboratory phases of the studies. This number is suggested, since it will provide good power (beta»alpha-0.05, 1 tail) to detect 2-fold increases in the relative risk for health outcomes which ordinarily occur at the rate of 1.5-2.0% (see Table 2). A major limitation of the sample size calculations for the cohort studies is that no good data exist on the expected prevalences of the outcomes postulated to be associated with TCDD exposure (see Table 3) in populations similar to the veterans to be studied. The occurrence of many of these conditions has never been assessed in population-based surveys. For some conditions there are data for men of the relevant ages from NCHS's Health Interview Survey (HIS) and Health and Nutrition Examination Survey (HANES). However, these national surveys may not accurately estimate the rate of chronic diseases in veterans — men who had to pass fairly rigorous medical examinations to get into the Army. In a sense, we will not be certain of the actual statistical power to detect increases in specific diseases until the analysis is under way and we know the frequency of the specific diseases in the unexposed cohorts. Berhaps this discussion begs the question: How were the sample sizes for each cohort of 6,000 for mortality assessment and interview and 2,000 for examination and laboratory testing chosen? Because of the paucity of relevant prevalence data, these choices were necessarily somewhat arbitrary; however, CDC believes that they are appropriate to detect an increased risk of important health outcomes in exposed veterans. For example, on the basis of data from the SEER network the cumulative total cancer incidence in the "unexposed" groups of veterans from 1968 to the time of the interviews is expected to be about 6 per 1,000. Therefore, we will be able to detect a 2-fold increased risk for this critical outcome (and all outcomes that occur in more than 5 per 1,000 of the unexposed). For the examination and laboratory testing phases we should be able to detect 2-fold increased risks of abnormal outcomes for dichotomous variables that occur in more than 1.5% 2.0% of the unexposed. On the basis of HIS and HANES data, these should include such important conditions as ischemic heart disease and diabetes �Page 27 mellitus. For continuous outcome variables, such as the results of most laboratory tests, we should be able to detect even modest differences between the exposed and unexposed groups. The power calculations have been made on the assumption that categorical data analysis will be done on the basis of a single 2 x 2 table for each disease. It is very unlikely that the situation will be simple enough to allow such straightforward analysis. Rather, it is anticipated that analysis will involve multiple variables (see section 4 6 2 ) and if unnecessary ... variables are inadvertently included, this may reduce power. Although the reduction should not be great, the situation is far too complex to allow any a priori estimation of just how large it may be. Another factor that may reduce power is misclassification on the variables used to define the cohorts ("exposure" variables) — if the misclassification is random. Of particular concern is the possibility that the records that have to be used to define the first two Agent Orange study cohorts ("likely exposed" and "likely not exposed") are so incomplete and/or inaccurate that there will be a sizeable amount of random misclassification in respect of true herbicide exposure. If this is the case, then power will be reduced, possibly to a significant degree, and the measures of effect will be biased toward the null. If misclassification in respect to exposure is present and not random, power would also be affected, and the measures of effect could be biased toward or away from the null. To achieve the power desired in the interview phase, it will be necessary to begin with cohorts larger than 6,000 because some of the desired study participants will not be located and some, once located, will decline to participate. CDC recommends that the goal for this phase should be a location rate of 85% and an 85% interview rate among those located, for an overall participation rate of 72%. Therefore, CDC recommends that the AAOTF select 8,350 (approximately 6,000/0.72) veterans for each of the cohorts. If the interview phase is successful, it should not be difficult to obtain the cooperation of 2,000 men per cohort for the examination phase, since there will be a pool of 6,000 to draw from. However, there is considerable concern that we may have difficulty in achieving a high rate of participation among those who are selected for inclusion in this phase. In other words, our concern here is not that we will be unable to reach the desired sample size of 2,000 per cohort but rather that participation might be limited to a highly selected group of men. We believe that the best we can hope for is a rate of 60% cooperation (i.e., 83% of the subsample composed of those who are located and agree to be interviewed [ . 3 0 6 / . 2 ) This may 08=.007]. be an optimistic goal. The Ranch Hand study team had an examination-phase participation of 87% among the Ranch Banders and 76% among the controls. As noted in section 4.2., CDC believes that the Air Force success can only be a goal which we can hope to emulate but not necessarily achieve. The NCHS experience of about 70% participation in its Health and ftitrition Examination Surveys can also be considered (the interview survey cooperation was about 95%). CDC believes that inferring directly from this experience to its own situation probably gives a somewhat optimistic expectation. The NCHS examinations were done in trailers located within easy commuting distance of the study participants, whereas most of CDC's study subjects will have to be transported to the examination sites by air (see section 4 3 1 2 ) Moreover, ..... the NCHS sample included persons of both sexes and all ages, whereas CDC's �Page 28 cohorts will be composed wholly of men of a narrow age range, a group that will probably have a lower-than-average propensity to participate. It will be desirable to assess study participants and non-participants with respect to differences in health and differences in exposures to health-influencing factors. Some assessment of this sort will be possible for the examination phase—men who are interviewed and who are invited but decline to participate in the exams will be compared to men who are examined. This comparison will make use of data gathered in the interviews. Unfortunately, a similar type of comparison cannot be made for those who are interviewed and those who are not. CDC will have very little, if any, health-related information about men who will not participate or who are not located. If feasible, comparisons will be made between interview respondents who readily participate and those who agree to be interviewed only after considerable coaxing. Similar comparisons could be made between veterans who are easy to locate and those traced only with considerable difficulty. Although not ideal, such comparisons may provide insights into the characteristics of those refusing to participate and those not located. 4 4 2 Selected Cancers Case-Control Study ... As with the cohort studies, the power of this study to detect a real increased risk among Vietnam veterans will depend on several factors, in this instance the number of cases and controls interviewed, the proportion of controls who served in Vietnam (and/or the proportion exposed to herbicides), the amount of exposure misclassification (misclasslfication of disease should be held to a minimum through the use of panels of pathologists, section 4 1 3 ) and the magnitude of the increased risk. To maximize the possibility ..., for including veterans who could have been exposed to Agent Orange in Vietnam, the study sample will include only men born from 1929 through 1953. Men born during these years ranged from 18 to 35 years of age during the time of maximum U.S. involvement in Vietnam, 1964 through 1971. Not including men with birth years before 1929 is expected to result primarily in the exclusion of non-combatant commissioned and non-commissioned officers, veterans who can be presumed to have had a low likelihood of exposure. By using VA data, the overall prevalence of service among men who will be 30-54 years of age in 1986 in the SEER areas has been estimated as 7.4% (Table 4). Power calculations for a 2-fold increase in risk among Vietnam veterans in general are presented in Table 5. Ages 30-54 are chosen as a reasonable approximation to the ages of men born 1929-1953. We have decided to study about 1,300 controls (i.e., equal to the projected numbers of lymphoma cases), since this number will give fairly good sensitivity for a 2-fold increase in risk for Vietnam veterans in general and since adding further numbers to the control sample will do little in terms of improving the power. The computation of power to enable the detection of a 2-fold increase in risk for Vietnam veterans in general requires explication. The Swedish studies which have suggested an association between herbicide exposure and sarcomas found risk increases of about 5-7. Thus, it would be reasonable to base power calculations for this study on relative risks of this magnitude and on an estimated prevalence of "meaningful" exposure among Vietnam veterans. As discussed elsewhere in this protocol, the records available for exposure estimation are not sufficient to allow a determination of "meaningful" �Page 29 exposure. Therefore, CDC has designed the study to be powerful enough to detect a generalized increase of 2-fold among all Vietnam veterans. However, Table 5 also includes power calculations made on assumptions that various fractions of Vietnam veterans might have been exposed. The power of the Selected Cancers Study for lymphomas will be higher than .for sarcomas because the number of cases is larger. Likewise, the power to detect increases in risk for liver, nasal, and nasopharyngeal cancers will be lower because of smaller numbers. It is unlikely that small real increases in risk can be demonstrated, even for lymphomas. Moreover, if Agent Orange or some other factor really has increased the risk of exposed veterans a small amount, and if only a small proportion of veterans were exposed to a toxic dose, the sensitivity of this study will be much lower than the figures presented. This will be a large case-control study, based on all soft tissue sarcoma and lymphoma cases that have occurred in a population of about 3,769,000 males aged 30-54 over a period of 4 years. Viewed from a somewhat different perspective, it will have roughly the same sensitivity as a very, very large cohort study, the cost of which would far exceed the cost of the proposed study. 4.5. Pretests and Pilot Studies 4.5.1. Agent Orange and Vietnam Experience Studies Two major categories of procedures need to be assessed before the main studies begin. First, there are a number of issues involving the manipulation of military records which need more work. Second, there is the matter of locating study subjects, securing their cooperation, and assessing the various study instruments (questionnaires, examination and laboratory protocols). The failure of any of the proposed procedures in preliminary tests will require revision of the procedures, and, if major failures are identified, outside consultation and peer review of new proposals. All proposed study procedures will be tested in a series of interrelated pilot studies and pretests. For the purpose of the discussion here, the term "pilot" study will be reserved to refer to the final process of assessing participation rates and evaluating interview and examination instruments just before the start of the main cohort studies. The term "pretest" will be used to refer to evaluations of all other procedures. It might be desirable to do formal and complete pilot studies for each of the three proposed studies. However, because such an approach would unnecessarily lengthen the time required to complete the two cohort studies, CDC recommends that procedures be tested with a series of related "pretests" and "pilot" studies. In those situations where one among several alternative procedures clearly seems to be the method of choice, only that method will be pretested and the other alternatives will be tried only if the preferred choice fails. In other instances* there may be no clear preference and then more than one procedure will be pretested. The general approach for the pretests will be early and close monitoring of circumscribed aspects of the study procedures. Several pretests of procedures which would be sequentially applied in the main studies can be done simultaneously. It is obvious that much time could be saved by using this approach. On the other hand, if problems are identified, there would be minimum delay, and relatively little work would be necessary to repeat the �Page 30 process with corrected procedures. Moreover, if no major problems are identified, then the data generated during the pretest could be used for the next pretest step or, for some procedures, the processes judged to be successful in pretests could be used straightaway for the main studies. An example of the pretest approach is the evaluation being done now to assess the locatability of male veterans and the plans for making the same sort of evaluation for female veterans. The AAOTF has transmitted to CDC identifying information for some 840 male veterans, and CDC has sent the information to the IRS to begin the locating process. The veterans used for this pretest were chosen because they were attached to two units that the AAOTF had worked with previously (1st of the 9th and the 31st Engineers). The AAOTF had the names of the individuals who served in these units in 1967-1968 at hand and only needed to request the personnel records from the St. Louis records center in order to obtain such items as SSNs and names and addresses of relatives. If the result of the locating pretest on this sample is encouraging, CDC will believe that the locating process does not need to be tested further before it embarks on the pilot study (see below). On the other hand, if the result is clearly discouraging, then CDC might recommend another study approach (see section 4 2 ) In either case, time could be saved and ... delays in reporting results to veterans held to a minimum. 4.5.1.1. Military Records Pretests Because AAOTF has had extensive experience in working with records from the Vietnam era, it is not expected that major problems will be discovered in the area of records manipulation. Even so, a more comprehensive test of the proposal to derive a sample of men for the Vietnam Experience study from the St. Louis records center seems in order, particularly to evaluate any problems that might arise in attempting to make the non-Vietnam veteran cohort match the Vietnam cohort in regard to calendar years of service (see section 4 1 2 ) To this end, a pretest sample of 200 Vietnam veterans and 200 .... non-Vietnam veterans will be chosen. If serious problems are identified with the procedures, then the process will be repeated with corrected procedures. The samples of veterans gathered during the (ultimately) problem-free pretest will be used as a part of the pilot study (see below). Much work needs to be done with the records that will be used to classify exposure. Although abstracting such data as daily unit locations is apparently simple, at least for those familiar with the records, so little actual work in this regard has been done for the purpose of assessing herbicide exposure that it must be considered a relatively untried process. Rather than incorporate this phase into a formal pilot study, it is proposed that the process be evaluated by constant monitoring during the preliminary unit selection process when the locations of the 50 battalions are identified on the randomly chosen days (see section 4 1 1 ) Even less experience has .... been accrued in the process of checking troop locations against the herbicide records. In particular, the schemes proposed in this protocol for scoring herbicide encounters have not been tried and their usefulness is unknown. Two pretests of these schemes will be made. The first pretest will take place when the randomly selected units from III Corps are evaluated for the purpose of ranking them on the herbicide encounter scores (section 4.1.1.); if there appear to be no problems at this stage, then CDC will have the AAOTF immediately proceed to the next step of the study, which will be the choice of �individuals for the main studies. tested again for individuals. Page 31 Later, the encounter scoring scheme will be 4.5.1.2. Location Rate, Participation Rate, and Instrument Assessments As mentioned above, some parts of the evaluation of the locatability of the cohort study subjects are now under way. This will continue as a part of the pilot study. Besides providing more information about locatability, the cohort pilot study will give information about expected main study participation rates and about possible difficulties with the interview instrument and examination protocol. The pilot study will be nearly a main study in miniature, the major exception being that the proposed selection process for the Agent Orange study cohorts will not be used to choose any of the pilot study subjects. As mentioned above, the subject selection process for the Vietnam Experience study will provide 400 veterans for the pilot study. Rather than wait for the process of ranking the companies in the 50 battalions from 111 Corps to be completed before selecting a pilot sample for the Agent Orange study, CDC recommends another approach to save time. It is proposed to simulate the Agent Orange main study through the use of 400 veterans who will be chosen from among the 110-120 combat battalions stationed in III Corps during 1967-1968. The selection of these pilot study veterans will involve the initial random selection of 10 companies from the 110-120 battalions. From each of these companies, 40 randomly chosen men will be selected. Although the cohort pilot study will simulate the main studies, the results will be considered in two stages — an interview stage, which will almost certainly be completed first, and an examination stage. If the interview stage proves to be successful, CDC will proceed with the interviews for the full study samples, even though the results of the examination stage may not be available. As noted elsewhere, CDC is concerned that it may be difficult to reach an acceptable level of participation in the examination phases of the studies. The Ranch Hand study group's enviable success in this regard is attributed in large measure to their treatment of study participants as "VIPs." CDC will attempt to duplicate this treatment. Since monetary factors may influence participation in the examination phase, CDC will test the effect of recompensing the subjects for lost wages; offering recompense may help to raise participation or if the offer offends a sense of altruism, it may decrease it. In addition, the effect of travel to distant locations for the examinations may enhance or deter participation. If it appears that more than one examining center will need to be used in the main studies (see section 4 3 1 2 ) the effect of distance to the center will be tested in the pilot ...., studies. 4 5 2 Selected Cancers Case-Control Study ... The Selected Cancers case-control study will be given a full pilot study in 2-3 SEER centers, each using 10 cases of lymphoma and 20 controls. Only lymphoma cases will be used because of the rarity of cases of the other cancers, and CDC cannot risk "wasting" them on a pilot study. Only 2-3 SEER centers will be used to minimize the time required — CDC believes that more are not required because of its previous success with the Cancer and Steroid �Page 32 Hormone study. The main purpose of a pilot study will be to evaluate the participation rate of males aged 30-49 and the interview instrument (CATI will not be developed for this pilot study, see section 4.3.1.2.). The work done by the AAOTF on scoring herbicide exposure likelihood for CDC's birth defects study (section 4.3.2.) is considered a valid surrogate for an assessment that could be done specifically for this study. 4.6. Data Analysisand Quality Control 4.6.1. Timing of Analyses The preferred approach to the timing of the analyses and the release of findings from the cohort studies is not easily found. Veterans will have considerable interest in receiving information about study results as soon as possible, and this suggests early analysis and release of significant findings even while data are being collected. But there are dangers in this approach. Locating individuals for the cohort studies can take considerable time, and, therefore, the early participants will be those who are easy to locate. One may speculate that the health of those who are easy to find differs from those who are difficult to find. If this is so, then early analysis could give a misleading picture and, ultimately, release of such results could be damaging. Although this consideration is cause for reluctance to make early analyses, it is also desirable to keep open the option of changing the interview instrument and examination procedures to accommodate some relationships noted in early interviews and examinations. In effect, the study itself could be used to generate hypotheses as well as test them. Having the flexibility to add procedures or questions to the examinations and interviews would also make it possible to accommodate new hypotheses which derive from sources outside these studies (examples of such outside sources include the VA's Agent Orange Registry, the Ranch Hand study, CDC's study of people exposed to dioxin at Times Beach, Missouri, the Australian studies of veterans, and the studies of U.S. Vietnam veterans being conducted by several state health departments). Given the lack of strong hypotheses at the outset, this is attractive. Biases could result from changing procedures if the changed procedures were disproportionately applied to difficult-to-locate individuals. To avoid this problem, CDC will divide the study subjects into groups for release for location and interview on a monthly basis. Changed procedures will only be used for those groups that have not yet been released at the time the changes are made. On balance, CDC believes that it is best to do analysis on a regular basis as the data are collected and to use the results to amplify or correct the thrust of the investigation. No findings will be released before all data collection and analysis is complete for some particular study phase, unless CDC, in consultation with its steering committee (section 9.), determines that it is mandatory that the preliminary analyses be released. An example of a finding which could not be withheld would be a convincingly substantial increase in the risk for a serious disease, especially if there are possibilities for effective treatment if the malady is diagnosed in its early stages. �Page 33 The concern about possible differences between study subjects who enter the cohort studies early and late does not apply to the Selected Cancers study. Therefore, CDC does not have the same level of concern about early release of findings from the case-control study. However, early findings which are released and later modified by further data collection will be difficult for the public to understand. On balance, CDC recommends the same approach as suggested above for the cohort studies. 4.6.2. Summary of Analytical Approach The two types of studies use somewhat different philosophical and analytical approaches to reach the same end, viz., the comparison of the risk of contracting certain diseases in those exposed to herbicides (and/or Vietnam service) with those not exposed. The two cohort studies provide direct estimates of disease incidence or prevalence, since the studies will begin with men who are selected because of some "exposure." Case-control studies usually do not provide estimates of disease rates or risks. However, the Selected Cancers study, being a population-based case-control study, will provide some insight into the incidence of the specified cancers among Vietnam veterans and among other men. This statement should not be taken to imply that this approach is equivalent to a cohort study, since the base population data is estimated by a random digit dialing census and could be influenced by incompleteness of the census because of lack of telephones and by migration. It is anticipated that a major part of the analyses will focus on the association between the presence or absence of disease and Vietnam service and herbicide exposure. For this part of the analysis, the primary measure of association will be the odds ratio, and the analytical techniques used will be those appropriate for dependent variables that are categorical. Other analyses will focus on dependent variables that are continuous and more appropriately dealt with by such techniques as the analysis of variance (Scheffe, 1959; Anderson, 1958) or non-parametric analogues (Puri and Sen, 1971). For example, a traditional approach to the data to be derived from some of the psychological tests would be to use multivariate analysis of variance as the primary analytical tool. For the sake of brevity, categorical data analysis is emphasized in the description that follows. However, it is to be noted that different but analogous techniques will be used for analyses involving continuous dependent variables. It is desirable that the measures of association (e.g., odds ratios) should be as free of the effects of other variables as possible; in other words, the estimates should be free of confounding effects. Therefore, the initial phases of analysis will be a search for factors that confound the estimates of association. This is not a simple matter. A primitive way to approach the problem is to compare (for a specific health outcome, exposure status and potential confounding variable) the crude odds ratio with the odds ratio adjusted for the potential confounder. If the two odds ratios are substantially the same, then the variable is not a confounder, at least within the study data, and need not be considered further. If it is determined that adjusting for the variable does alter the odds ratio in the data at hand, then it must next be determined if the variable independently predicts disease and exposure. If it does independently predict, then the variable will be included in further �Page 34 analyses. If, on the other hand, the prediction is not independent, then the variable may be a part of the causal chain and it should not be used as an adjusting variable. To illustrate, suppose we consider education as a potentially confounding variable in one of the cohort studies. The first step would be to determine if adjusting or "controlling" for education changes the odds ratio substantially. If not, then education can be ignored in further analysis of the specific disease-exposure relationship. If adjusting for education does substantially alter the odds ratio, then it will be determined if education is related to disease within the "exposed" and "unexposed" groups, that is, it will be determined if education predicts for disease independently of exposure status. If education is only related to disease through the agency of cohort status, or vice versa, then it may be omitted in further analysis. The flaw in this approach is that there may be other variables which modify the association between the variables being considered pairwise (i.e., in statistical jargon, higher order interactions). For example, education may be associated with memory of key factors which are, in turn, associated with disease and service. Thus, this primitive approach to discovering confounding variables has merit primarily because of the ease with which it may be accomplished and because it can be used for categories of disease with relatively small numbers (see also below). Under these circumstances, the final estimate of the effect measure for a particular classification of disease would be done by a method such as that of Mantel and Haenszel (1959). This procedure will yield a summary odds ratio and test statistic (or related confidence limits) for the several 2 x 2 tables (Vietnam Experience study example) Vietnam Service Yes No Yes a b Disease X c No d which have been formed on the basis of one or more confounding variables. A better (but not infallible) way to perform a detailed assessment of variables which influence the association between Vietnam service and cancer is to consider them in a multivariate framework. The analytic technique to be used will be log-linear analysis or a related technique, such as logistic regression or proportional hazards modelling (Bishop et al,, 1975; Breslow and Day, 1980; Cox, 1970). The basic approach can be illustrated by considering the simple case of a 2 x 2 x 2 table with race as the third variable of concern: White Vietnam Service Yes Yes a No b No c d Black Vietnam Service •> Yes No a b Disease X c d. �Page 35 It should first be determined whether the odds ratios In whites and blacks are substantially the same (i.e., does race modify the association between service and the disease). If the odds ratios are not substantially different then one need only consider the association between service and disease (with possible adjustment for confounding). If the odds ratios are substantially different, in whites and blacks, the association between service and disease should be considered separately for each race. In actuality, the problem will be much more complex. Many variables are potential confounders or modifiers of the association between various diseases and service, and, consequently, it will be necessary to consider numerous 2 x 2 tables. Although analysis by such methods as logistic regression is, in theory, well suited for this problem, difficulties will arise. Stratification over increasing numbers of variables rapidly produces so many 2 x 2 tables that there are no observations in many table cells. The method then begins to break down. We, therefore, have to make some compromise between the desired degree of stratification and search for confounding and higher order interactions and what will be practicable within the framework of these studies. In summary, we propose to do our analyses starting with the simple stratification techniques on relatively limited numbers of variables and, as we learn more about the data, we will progress to control of confounding and model building by the more ambitious logistic regression or related techniques. 4 6 3 Quality Control ... The success of the above methods of analysis in assessing the association of herbicide exposure and Vietnam service with adverse health outcomes is predicated on the accuracy of the data being analyzed. CDC has conducted many nationwide epidemiologic studies and is experienced in dealing with the important issues of quality control and data validation. Many of our approaches to these issues have already been mentioned. For the Agent Orange study CDC has requested that the National Academy of Sciences make a further assessment of the critical information on herbicide applications contained in the "Herbs" computer tape (see section 4 1 1 . For ..) both the Agent Orange and Vietnam Experience studies, we will attempt to achieve rigid quality control for both the laboratory testing and physical examinations (section 4.3.1.2.B) and the questionnaire administration. Central to the latter effort will be our use of computer-assisted telephone interviewing (CATI) (section 4.3.1.2.A). In addition, for the mortality analysis for these studies we will assess the extent of underascertainment of deaths for each of the cohorts (section 4 3 1 1 . ...) Among our quality control measures for the Selected Cancers study are an expert panel review of the histologic material used for diagnosing the cancer (section 4.1.3) and blinding both the CATI interviewers and the AAOTF personnel responsible for assessing Agent Orange exposure as to the case or control status of the study participants (section 4 3 2 . ..) In addition to these approaches, emphasis will be given to evaluation of non-participants (section 4 4 1 . Where feasible, we will attempt to verify a ..) sample of hospitalizations and participant-reported illnesses with the �Page 36 relevant health care providers. We will take special care to ensure standardization of methods if more than one examination and/or laboratory center is needed (section 4 3 1 2 B . These efforts will include evaluating ....) volunteers at more than one examination center to assess the between-center variability. CDC is committed to conducting the best possible assessment of the health of Vietnam veterans. We will make every effort to obtain the best quality information on the health of study participants. Where possible, we will assess the extent of any inaccuracies in our data. �Page 37 5. Inferences from Possible Study Findings; StudyLimitations A major concern of Vietnam veterans is that they are at high risk for quite a variety of diseases. The cause of this putative high risk is generally suspected to be exposure to Agent Orange and other herbicides, but there is also concern that other factors incidental to Vietnam service may have conferred an increased risk. The design of CDC's studies should permit an assessment of both general and some specific concerns. The Agent Orange study will permit an evaluation of the possible health consequences of herbicide exposure, and the Vietnam Experience study will give information regarding health risks that may be associated with the general (Army) service experience. Unavoidable limitations of the proposed studies, or indeed any other studies which could be done, will preclude describing the results as "definitive." A number of limitations have already been mentioned, but some of them need to be repeated here, and a few more need to be added. An important limitation is that the proposed studies are observational, as opposed to experimental, and observational studies inherently require some tempering of the inferences drawn from them. Another general caveat is that it is not possible to prove a negative — that is, it will never be possible to say with certainty that herbicide exposure or some other factor connected with Vietnam service did not cause any adverse health effects. In addition, when evaluating negative findings, the study power, or sensitivity, must always be kept in mind. The proposed studies will be quite powerful, but they will not provide answers to all health questions that might arise. However, if no increase in risks is found, these studies should be of substantial value in easing the concerns of veterans. The ability to detect such specific increases will depend on the magnitude of the risk and the numbers of veterans (cases and controls in the Selected Cancers study) studied; the possibilities for exposure misclassification between the "likely exposed" and "likely not exposed" cohorts in the Agent Orange study have already been mentioned as a cause of concern. Moreover, even in the absence of exposure misclassification, the studies will have low power for rare diseases and/or low increases in risk, or for increases in risk limited to those veterans with prolonged and/or heavy exposure to herbicides or some other harmful factor. Thus, an overall finding of no increase in risk might "hide" a real increase for specific disease categories or special groups of veterans. But if the increase is limited to very rare categories of disease or to special veterans, then the study still has the utility of putting some boundary on the scope of the problem for most veterans. The lack of strong hypotheses has been mentioned previously and this has led us to propose a rather wide ranging investigation. Thus, we may not give enough emphasis to some crucial factor. Our proposal to keep open the option of modifying our interviews and examinations mitigates this concern somewhat. However, it is conceivable that we will not include some critical item in our investigation, and from this type of omission there is no recovery. Depending on the results of analysis, the design of the Agent Orange study may present unusual problems of inference. Some examples follow. If the first cohort ("likely exposed") appears to have significantly higher �Page 38 disease risks than the second cohort ("likely not exposed") and the third cohort, then, depending on such considerations as the magnitude of the increase in risk, the inference will be clear — herbicide exposure confers a health decrement. But suppose that the first and second cohort have similar disease risks and that they are both higher than the third. Then, one will be at a loss to say if the lack of difference between the first two and their similar difference with the third is due to exposure misclassification in the first two cohorts or to the difference in service experience. Another problem of inference will be false positive findings. We plan to make comparisons of presumed herbicide exposure and/or Vietnam service for numerous health outcomes. There is, therefore, a certain probability that several of these will show statistically significant positive associations even if, in truth, there are none. It is difficult to a priori specify how these are to be handled. It may be that some such associations will be "convincing," in and of themselves, whereas others may not. Making such inferences transcends from the cold objectivity of statistics to the art of medicine — at this stage considerations such as the biological plausibility of associations play a large part. In addition, the following approach may help in making such judgments. If the number of significant associations found is reasonably close to the number expected under the null hypothesis (e.g., 5% significant if working at an alpha - 5% level) and if the associations are relatively well balanced with respect to the direction of the association (e.g., if the number of instances where presumed herbicide exposure and/or Vietnam service appears harmful is approximately the same as where service appears protective), then we might be inclined to attribute the significant findings to chance. Finally, it is not unlikely that we will be left with equivocal positive results. �Page 39 6. Report of Study Findings CDC will prepare comprehensive reports of the findings for each of the study phases. The credibility of the results will be enhanced if the major findings are released simultaneously in peer-reviewed medical journals. �Page 40 7. Timetable_, Milestones, and Reports Month 1 in the following timetable is December 1983. The timetable is ambitious and may be difficult to follow. CDC will do its utmost to ensure that there are no avoidable delays. It is projected that the Selected Cancers study will be finished last, at Month 69. The rate limiting factor for this study is the relatively low number of cases that will accrue each year. If CDC can identify other population-based^ cancer registries that have good case-ascertainment rates and that are willing to participate, the completion date would be sooner than the date currently projected. Month Number 1 4 7 9 10 11 12 13 14 16 17 18 23 29 30 33 39 42 45 49 52 58 63 69 Major Milestone - begin selecting Vietnam Experience (VE) main study subjects obtain OMB approval Random Digit Dialing Contract Award Selected Cancers (SC) Data Collection Agencies Contract Award - Agent Orange (AO)-VE Interview Contract Award - begin interviews, AO and VE pilot studies - SC interviews begin - SC Pathology Contract Awards - Examinations Contract Award(s) - Company location for first 25 battalions complete, AO study - VE study main interviews begin - assess AO and VE pilot study - begin VE study medical exams - begin selecting AO main study subjects - selection of VE study individuals complete - company location for second 25 battalions complete, AO study - complete VE study mortality data collection - report VE study mortality data - complete VE study interviews - complete VE study medical exams - report VE study interview data - report VE examination data - complete AO study interviews - report AO study mortality data - complete AO medical exams - report AO study interview data - report AO study exam data - complete SC study histological review - report SC study data �Page 41 8. Investigators These studies will be conducted under the direction of the Agent Orange Projects, an organizational entity located Diseases Division of CDC's Center for Environmental Health; laboratory work will be by the Clinical Chemistry Division, Center for Environmental Health. staff assigned to in the Chronic oversight of also of CDC's The following staff, drawn from CDC's Agent Orange Projects group and Cancer Branch, have contributed to the scientific development of this protocol: Lee Annest, PhD; Edward Brann, MD, MPH; Pamela Byrnes; Pierre Decoufle1, ScD; J. David Erickson, DDS, MPH, PhD; Nancy V. Hicks, RN, MS; Michael Kafrissen, MD, MPH; Peter M. Layde, MD, MSc; Maurice LeVois; Marion R. Nadel, PhD, MPH; Thomas K. Welty, MD; Matthew M. Zack, MD, MPH. Robert Diefenbach, John Gallagher, Peter McCumiskey, Melvin Ralston, and Joseph Smith have provided technical and administrative support, and secretarial assistance has been given by Gerri Culpepper, Teresa Ellington, Janiece Myers, Emily Peters, Jean Reynolds, Hazel Riley, and Effie Spencer. The staff of the Army Agent Orange Task Force, under the direction of Richard C. Christian, has given valued advice. �Page 42 9. Protocol Review; Study Oversight A draft of this protocol received wide scientific review. A panel of CDC scientists from programs outside of the division responsible for the studies conducted a scientific evaluation. The Office of Technology Assessment, the Science Panel of the Agent Orange Working Group, and the Advisory Committee on Special Studies Relating to the Possible Long-Term Health Effects of Phenoxy Herbicides and Contaminants also conducted scientific reviews. In addition, CDC transmitted copies of the draft protocol to the representatives of about 15 veterans' organizations for their consideration. This version of the protocol incorporates a number of changes suggested during these reviews. The written reviews received, and CDC?s responses to them, are available on request. Since the detailed interview instruments and examination protocols are currently being developed, CDC will make these available on request to interested parties when they are completed. This version will receive "human subjects review" by CDC's Institutional Review Board and review by the Office of Management and Budget. CDC will conduct the studies with guidance from a steering committee. It has been requested that a subcommittee of the panel which provides oversight of the Ranch Hand studies be formed for this purpose. CDC proposes that steering committee meetings be held at 6-month intervals, to be supplemented by other meetings as the need arises. �Page 43 Table 1 Cumulative Expected Numbers of Deaths by Cause in a Hypothetical Cohort of 6,000 Men Aged 22 in 1968 and Followed Through 1984 (17 Years) Cause of death2 All causes Expected Number of Deaths 213.0 Accidents (E800-E949) Motor vehicle (E810-E823) Other (E800-E807, E825-E949) 79.1 48.3 30.8 Suicide (E950-E959) 25.5 Homicide (E960-E978) 27.3 Diseases of Heart (390-398, 402, 410-429) 18.6 Malignant Neoplasms (140-204) 17.3 Cirrhosis of liver (571) 6.6 Cerebrovascular diseases (3-3) 4048 3.6 Influenza and Pneumonia (470-474, 480-486) 2.9 Diabetes Mellitus (250) 2.1 Nephritis and nephrosis (580-584) 0.7 Bronchitis, emphysema and Asthma (490-493) 0.5 Septicemia ( 3 ) 08 0.5 All other causes (residual) 28.2 ^Expected numbers based on 1978 U.S. age-specific rates for males. The age-specific rates were quinquennial (5 years), and the cumulative rates used to derive the expected numbers were computed by weighting the quinquennial rates by the number of years of cohort experience in each quinquennium (constant cohort size). Source of rates: Vital statistics of the U.S.:1978, Vol. II, Mortality Part A, NCHS, 1982. ^Numbers in parentheses are the relevant codes from the Eighth Revision International Classification of Diseases, Adapted. �Page 44 Table 2 Power1 to Detect Various Relative Risks in the Agent Orange and Vietnam Experience Studies, by Prevalence of Condition in "Unexposed" Group A. Interview Phase (6,000 per group) Prevalence per 100 of Condition in ^Unexposed" Group 2 Relative Risk 4 6 0.10 0.321 0.928 0.998 0.20 0.576 0.998 0.999+ 0.30 0.750 0.999+ 0.35 0.811 0.40 0.859 0.50 0.923 1.00 0.997 1.50 8 0.999+ 0.999+ Power calculations with 1-tail, alpha - 0.05 by method of Casagrande JT, Pike MC: An improved approximate formula for calculating sample sizes for comparing two binomial distributions. Biometrics 1978;34:483-6. �Page 45 Table 2 (continued) Power1 to Detect Various Relative Risks in the Agent Orange and Vietnam Experience Studies, by Prevalence of Condition in "Unexposed" Group B. Examination Phase (2,000 per group) Prevalence per 100 of Condition in " UnexposedN Group 2 Relative Risk 4 6 8 0.10 0.475 0.778 0.923 0.20 0.218 0.794 0.975 0.998 0.30 0.321 0.930 0.998 0.999+ 0.35 0.370 0.960 0.999 0.40 0.416 0.978 0.999+ 0.50 0.502 0.994 1.00 0.796 0.999+ 1.50 0.926 2.00 0.976 2.50 0.993 3.00 1 0.108 0.998 Power calculations with 1-tail, alpha - 0.05 by method of Casagrande JT, Pike MC: An Improved approximate formula for calculating sample sizes for comparing two binomial distributions. Biometrics 1978;34:483-6. �Page 46 Table 3 Selected Health Outcomes Reported To Be Associated with Exposure to TCDD - Animal and Human Literature* Dermatologic ~"Chloracne Hirsutism Hyperpigmentation Hepatic IPorphyria cutanea tarda Hepatomegaly Elevated serum levels of hepatic enzymes Neuropsychologic Peripheral neuropathy Asthenia and lethargy Immunologic Impaired cutaneous delayed hypersensitivity response Increased risk of infection Reproductive Reduced fecundity Adverse pregnancy outcomes Cancer Soft tissue sarcoma, lymphoma, and nasopharyngeal and nasal General Lipid metabolism: Hypercholesterolemla and hypertriglyceridemia *This table is by no means an exhaustive list (see Appendix B for literature review). It is intended to show the wide range of health outcomes postulated to be linked to TCDO exposure. �Page 47 Table 4 Estimated Prevalence of Vietnam Service and Expected Number of Cases of Cancer for the Selected Cancers Case-Control Study In Males Aged 30-54 In 1986 in the SEER Areas Estimated Yearly Number of Cases3 MalesJ Age Prevalence of Vietnam Service2 Soft Tissue4 Sarcoma Lymphoma^ Primary Nasal and 6 Nasopharyngeal Liver 20 53 4 3 11.7 14 45 5 3 740 12.5 17 52 6 5 45-49 590 3.7 22 75 10 12 50-54 552 1.5 33 106 17 20 Total 3,769 7.4 106 331 42 43 30-34 980 35-39 907 40-44 4.9 Estimated number of males (thousands) in SEER areas, 1976 data projected to 1986, National Cancer Institute Monograph 57, 1981. Percent of males who are Vietnam veterans; estimated from VA data on numbers of Vietnam era veterans and assumption that 32.2% of Vietnam era veterans served in Vietnam. Incidence of cancers derived from National Cancer Institute Monograph 57, 1981. Includes the following (morphology-based) tumor types: fibrosarcoma, malignant fibrous histiocytoma, liposarcoma, leiomyosarcoma, rhabdomyosarcoma, Kaposi's sarcoma (estimate based on pre-AIDS incidence), blood vessel sarcoma, nerve sheath sarcoma, synovial sarcoma, malignant mesenchymoma, malignant paragarglioma. Incidence estimates also based on categories "sarcoma NOS" and "other sarcoma." Includes Hodgkin's Disease and non-Hodgkin's lymphoma. Includes the following topographic tumor types: nasopharynx, nasal cavity, accessory sinuses. Includes liver and intrahepatic bile ducts. �Page 48 Table 5 Power1 of Selected Cancers Case-Control Study to Detect Increased Relative Risks a) 2-fold Increase in Relative Risk for Vietnam Veterans in General Study Year 1 Type of Participant Soft Tissue Sarcoma Lymphoma Nasal & Nasopharyngeal liver Controls Number Control Group Prevalence of Vietnam Veterans 0.075 0.100 0.050 106 331 42 42 325 0.45 0.67 0.30 0.30 0.57 0.82 0.37 0.37 0.66 0.90 0.43 0.43 Study Year 2 Number2 Soft Tissue Sarcoma Lymphoma Nasal & Nasopharyngeal Liver Controls Control Group Prevalence of Vietnam Veterans 0.050 0.075 0.100 212 662 85 85 650 0,70 0,92 0.47 0.47 0.83 0.98 0.58 0.58 0.90 0.99+ 0.66 0.66 Study Year 4 Number2 Soft Tissue Sarcoma Lymphoma Nasal & Nasopharyngeal Liver Controls Control Group Prevalence of Vietnam Veterans 0.075 0.100 0.050 319 993 128 128 975 0.84 0.98 0.60 0.60 0.94 0.99+ 0.73 0.73 0.97 0.99+ 0.81 0.81 Study Year 4 Number2 Soft Tissue Sarcoma Lymphoma Nasal & Nasopharyngeal Liver Controls 425 1,324 170 170 1,300 Control Group Prevalence of Vietnam Veterans 0.050 0.075 0. 100 0.92 0.99+ 0.70 0.70 0.98 0.99+ 0.82 0.82 0.99+ 0.99+ 0.89 0.89 �Page 49 Table 5 (continued) b) 2-fold and 5-fold Increases in Relative Risk Under Assumption of 7.5% Control Group Prevalence of Vietnam Service and 3 Levels of Possible Agent Orange Exposure Among Vietnam Veterans (Study Year 4 Only) 2-fold Increase in Relative Risk For Agent Orange Exposed Vietnam Veterans Possible Prevalence of Agent Orange Exposure Among Vietnam veterans of Participant Soft Tissue Sarcoma Lymphoma Nasal & Nasopharyngeal Liver Controls Number'* 0.10 0.25 0.50 425 1,324 170 170 1,300 0.33 0.49 0.23 0.23 0.62 0.85 0.41 0.41 0.85 0.99 0.61 0.61 5-fold Increase in Relative Risk for Agent Orange Exposed Vietnam Veterans Type of Participant Soft Tissue Sarcoma Lymphoma Nasal & Nasopharyngeal Uver Controls Number^ 425 1,324 170 170 1,300 Possible Prevalence of Agent Orange Exposure Among Vietnam Veterans 0.10 0.25 0.50 0.96 0.99+ 0.81 0.81 0.99+ 0.99+ 0.98 0.98 0.99+ 0.99+ 0.99+ 0.99+ Power calculations with 1-tail, alpha - 0.05 by method of Casagrande JT, Pike MC: An improved approximate formula for calculating sample sizes for comparing two binomial distributions. Biometrics 1978;34:483-6. Estimated number of participants �Page 50 APPENDIX A (November 1982) Protocol Outline Tentative Timetable Epidemiological Studies of the Health of Vietnam-Era Veterans^ (Agent Orange) Overall Design The Centers for Disease Control (CDC) recommends two complementary historical or retrospective cohort studies. One study will compare the health of a group of U.S. veterans of the Vietnam conflict with the health of a group of Vietnam-era veterans who did not serve in Vietnam; it may Include individuals from all four branches of the military. The purpose of this study will be to make an assessment of the possible health effects of the general Vietnam service experience. The other study, which is designed to evaluate the health effects of possible exposure to herbicide Agent Orange, will compare the health of three groups or cohorts of Vietnam veterans who differ in their probable level of exposure to Agent Orange. This second study will focus primarily on veterans of the Army but will probably include veterans of the Marine Corps. Each of these two studies will have three major components: 1) a mortality assessment (mortality followup will be repeated every 5 years for the foreseeable future); 2) a health and exposure questionnaire; and 3) a clinical and laboratory assessment. The studies will have several other features in common. However, the sampling plans and some of the health outcomes measured in the questionnaire and clinical assessments will differ between the two studies. Moreover, they will follow different timetables. They are designed to answer related but distinct questions of importance to Vietnam veterans and their families. These two studies should be sufficient to meet the directive of Congress which instructed the Veterans Administration to conduct an "epidemiological study"; in addition, they are responsive to current veterans' and congressional concern. However, these studies are but a part of the Federal effort to provide answers about the possible health effects of herbicides and their contaminants, and about the effects of military service in Vietnam. Other major Federal activities include: 1) CDC's ongoing study which is designed to determine if Vietnam veterans are at increased risk of fathering babies with birth defects; 2) CDC's NIOSH Dioxin Registry, which will assess the health effects of occupational exposure to dioxin during the manufacture of herbicides and related chemicals; 3) the U.S. Air Force's comprehensive health study of veterans who applied herbicides in Vietnam from fixed-wing aircraft ("Ranch Hand" study); 4) the Veterans Administration's (VA) proportionate mortality study of Vietnam veterans; the VA is also supporting protocol development for a study of twins, one of whom went to Vietnam and one of whom did not. Composition of Cohorts and Sampling Plans The choice of individuals for inclusion in the various study cohorts will derive from review of military records from the Vietnam era. Considerable thought about and work with records from Vietnam has been done by the �Page 51 Department of Defense (primarily staff of the Army Agent Orange Task Force— AAOTF), the Veterans Administration, and the White House Agent Orange Working Group. A consensus seems to have been reached that the choice of individual veterans for an Agent Orange study will involve the use of personnel records and company level action records and a variety of herbicide usage records. More thought needs to given to the specific organization and analyses of records which might be used for a Vietnam Experience study, but it is recommended that company level records also be used for this study. a) Agent Orange Study A good design for a historical cohort study of the possible health effects of Agent Orange would involve the use of 2 groups of men who were as similar as possible in all respects except for their exposure to the herbicide* One group would ideally be free from all exposure while the others would have been subjected .to "meaningful" exposure. (Other attractive designs might include subdivisions of those exposed based on levels and/or duration of exposure, or even continuous measures of exposure for individual veterans.) It appears that such an ideal is not attainable. Obstacles include: 1) the military records which must be used were made during a war and, therefore, of uneven quality; 2) an inability to define objectively "meaningful" exposure; 3) the difficulty in ensuring that veterans who were possibly or likely exposed (by whatever measure) are comparable (with respect to all things which might influence health) to veterans who were not exposed. Under ordinary circumstances, such obstacles would probably prevent the initiation of an Agent Orange study. It is, therefore, mandatory that advance advice and consent be obtained from veterans' groups with respect to study policies and procedures, especially those directed at defining Agent Orange exposure. The important company records which give information about troops are the morning reports and the journal files. The morning reports can be used to document the presence or absence of individual servicemen on a daily basis while the daily journal files will indicate the locations of companies in time and space. The major herbicide records are those which document the time and location of fixed-wing aircraft applications of herbicide (Ranch Hand missions—contained on the "Herbs" tape), base perimeter applications records, and information about Ranch Hand mission aborts (dumps). The choice of an individual for inclusion in the "likely-exposed" cohort will be based on a measure of company proximity in time and space to herbicide applications as documented by these records. Members of the "non-exposed" cohort will likewise be chosen because of a measure of their company's distance in time and space from any herbicide applications. The company records may contain gaps ( . . whole periods of time missing) ie, and are probably quite variable in terms of quality and detail, because they were created during the war. The herbicide usage records are known to contain errors with respect to the time and location of applications and the degree of their completeness is unknown. They are far from ideal �Page 52 as the starting point for an historical cohort study. There may be opportunities to assess the accuracy and completeness of the herbicide usage records, and every effort will be made to pursue these opportunities. However, there are no possibilities for similar checking of the company troop records. Thus, the categorization of individuals with respect to their potential for herbicide exposure will be uncertain and will forever remain so. The desire to ensure that troops classified as "exposed" to Agent Orange are comparable to "non-exposed" troops with respect to other factors which might influence health is another issue which makes it difficult to design an "ideal" study. The underlying problem is that the use of herbicide was not equally distributed in Vietnam. Areas where it was heavily used were generally combat areas and differed in terrain and flora from those areas where it was little used. These areas may also have differed in other important respects, such as, indigenous diseases, level of combat intensity, and type of personnel deployed. It is for these reasons that much of the recent thinking about the subdivision of troops into "exposed" and "non-exposed" groups has been directed at choosing the cohorts from the same area of Vietnam. Unfortunately, because of the inherent limitations of the records, this approach may have the effect of increasing exposure misclassification (especially the categorization of those who are truly "exposed" into the "non-exposed" group). These two competing forces, the desires for comparability and for maximum exposure separation, have drawn CDC to recommend a three-cohort design. Two of the three cohorts will be from the same area of Vietnam (and time during the war) but will differ in regard to their exposure likelihood. These two cohorts will be comparable but suffer from imprecision of exposure separation. The third cohort will be drawn from another area of Vietnam (but from the same time period), an area where there is good evidence of little or no herbicide usage. This cohort will give maximum exposure separation from the "exposed" cohort but may suffer from a lack of comparability in respect of other health-influencing factors. This design is incomplete, as is illustrated in the following 2 x 2 table which cross-classifies exposure by a measure of general experience, which will be called "combat." Agent Orange Exposure Yes No Yes Cohort I Cohort 2 "Combat" No Cohort 3 The empty cell, representing the combination of Agent Orange exposure with no "combat," cannot be filled, because it is our understanding from the military that Agent Orange use was inextricably entwined with a certain "combat" experience. Because of its incompleteness, this design will present problems in analysis and interpretation. Moreover, the comparison of the first and third cohorts, which will ensure maximum exposure separation, may be subject to respondent bias; respondent bias should not be a problem in a comparison of cohorts 1 and 2, because individual respondents will probably be �Page 53 uncertain about their (study) exposure status. Despite these problems, we believe that this design is better than either of the other alternatives based on an approach which uses only two cohorts—either decreasing exposure misclassification by decreasing comparability or increasing exposure misclassification by increasing comparability. The results of the Ranch Hand study, currently being conducted by the U.S. Air Force, may help in the interpretation of this incomplete design. The Ranch Hand study will compare the health of crews who flew the herbicide spray missions with air crews who did not fly spray missions. Thus, it will provide information about Agent Orange exposure in the absence of the general experience of ground troops. b) Vietnam Experience Study The idea of studying ill-health effects which might derive from the "general experience" of having been in Vietnam is at once attractive and unappealing. It is attractive because there may have been many factors which could have adversely affected those who served in Vietnam, In contrast to their counterparts who served elsewhere. And it is also plausible that Vietnam veterans who did not see active combat in Vietnam were subjected to health-influencing events that were not part of the experience of those who served elsewhere. Any study which focuses on Agent Orange alone will obviously not test such a plausible multlfactorial hypothesis. However, the multifactorial nature of this hypothesis makes the study of the "Vietnam experience" unappealing from the scientific point of view. The "experience" comprises many factors, many of which are unknown, poorly defined, or not quantifiable. Nevertheless, it is our opinion that this is an important question to the Vietnam veteran, and one which deserves as much attention as the issue of the possible effects of Agent Orange. Viewed in the broadest terms, the Vietnam "experience" could have influenced anyone who served there. It is, therefore, suggested that consideration be given to the inclusion of veterans of the Army, Navy, Marines, and, if possible, the Air Force (the records systems of the Air Force might make inclusion of that service's veterans very difficult). A major concern about the validity of making a comparison of Vietnam and non-Vietnam veterans derives from an undocumented suspicion that there may have been preexisting differences between the two groups in terms of health-influencing factors and behaviors* If such differences existed and if they applied to all veterans, then a valid study of the Vietnam "experience" would not be possible. However, military personnel with whom we have consulted do not feel that such factors would have existed for all Vietnam veterans. Specifically, it is their belief that being sent to Vietnam was a matter of the "luck of the draw" for those who were drafted or who were short-term enlistees. Serving in Vietnam, the U.S., in Europe, or elsewhere was, in their opinion, a matter which depended on occupational specialty and the operational needs of the various commands. Thus, �Page 54 any given serviceman was at risk of serving anywhere where there was a need for his occupational specialty. Choice of individuals for the two cohorts of this study should be made after a review of company and personnel files in much the same manner as will be done for the Agent Orange study. A simple random sample or a stratified random sample of Vietnam veterans and non-Vietnam veterans would probably be the method of choice but the filing of the available records probably makes this infeasible. Therefore, we recommend a cluster sampling of military units (much as will be done for the Agent Orange study) and a random sampling within clusters as the method for selecting members of each cohort. Sample Sizes It is recommended that each of the 5 cohorts (3 Agent Orange study and 2 Vietnam Experience) be composed of 6,000 servicemen. All of these individuals will be included in the mortality studies, and it is hoped that up to 90% of the surviving cohort members will be included in the questionnaire phase of the studies. (The results of the Ranch Hand study, better than 95% interview completion, give reason to set such an optimistic goal. If, however, the questionnaire pilot studies give indications of completion rates much under 70 or 75%, careful consideration should be given to not proceeding with the main studies.) The number of 6,000 for each cohort was chosen because comparisons between 2 groups of between 5,000 and 6,000 each will be able to detect (alpha « beta * 0.05, 1-tail) 2-fold increases in the relative risk for health outcomes which ordinarily occur at the rate of 0.5%, for example, all cancers (detecting associations for specific cancers would require truly massive cohorts—this problem is probably best approached through specific case-control studies). For the clinical and laboratory phases, it is suggested that random samples of 2,000 from each cohort be chosen. It is hoped that as many as 80% of those chosen will participate and, as with the questionnaire phases, if the pilot study shows rates much below the 70% level, it will be necessary to question the wisdom of proceeding with the main study phases. The number 2,000 was chosen because samples between 1,500 and 2,000 will give good power (alpha » beta - 0.05, 1-tail) to detect 2.5-fold increases in the risk of outcomes which usually occur at the rate of 1.0%. (The major health outcome categories from which the questionnaire and clinical laboratory phases will be developed during protocol design and review are listed in a later section "of" this outline!) Study Sequences Three phases are planned for each of the 2 studies and each phase will culminate in a separate report. The 3 reports will concern 1) mortality experience of the cohort members; this phase of the study will also give an indication of the proportion institutionalized, 2) the results of the health questionnaire, and 3) the results of the clinical and laboratory tests. It is anticipated that work will proceed first on the Vietnam Experience study because there will be less work involved in selecting the cohort members than there will be for the Agent Orange study. Within each study, ascertainment of �Page 55 vital status will be a part of the process of locating cohort members for the health questionnaire and clinical/laboratory phases. Thus, mortality analysis will be completed first; reports on the health questionnaire and clinical/laboratory analyses will follow later. Even though these studies are subdivided into phases, it is expected that at some point in time work will be proceeding simultaneously on both studies (see schedule, later in this outline)« The major steps which will be required to complete the two studies are (after full protocol design and approval and after pilot testing of procedures): 1) Selection of individual cohort members by the Army Agent Orange Task Force (AAOTF) For the Vietnam Experience study, identifying information about the cohort members will be transmitted to CDC immediately after selection. For the Agent Orange study much more work will be required of AAOTF personnel because of the need to review exposure information. Identifying information about cohort members for each study will arrive at CDC in small batches, possibly on a monthly basis, as they are selected. Therefore, the selection will be done in such a way that an appropriate balance of "exposed" and "non-exposed" for the Agent Orange study and of Vietnam and non-Vietnam veterans for the Vietnam Experience study are included in each batch. 2) Vital Status Determination and Location of Cohort Members As soon as a batch of information for study individuals is received, a check will be made against the Beneficiaries Identification and Records Location System (BIRLS) files and the National Death Index to try to ascertain those individuals who are deceased. For those who are found to be dead, collection of death certificates, pathology reports and other relevant material will ensue. Procedures to determine the location of those currently alive will begin simultaneous with the checks against the BIRLS and National Death Index—the first step will be to check against Internal Revenue Service (IRS) files, which is a rapid and inexpensive method to obtain relatively current addresses for taxpayers. For those individuals who are not found on the BIRLS file or National Death Index and who are also not found on the IRS files, more expensive and time consuming methods of location will be used. The goal for both studies will be a location rate of 95% for those who are presumed alive. 3) Health Questionnaire Interviews of about 45 minutes in length will be conducted by telephone where possible. For potential respondents without telephones, personal interviews will be conducted at a place convenient for the respondent; for potential respondents who are institutionalized, personal interviews will be conducted at the place of instltutionalization. The major outcomes from which questionnaire items will be chosen during the stage of full protocol development �Page 56 are listed later in this outlinet The goal for both studies will be an interview completion rate of better than 90% of those located. 4) Clinical and Laboratory Examinations Clinical examinations of the 2,000 individuals from each of the 5 cohorts will take place at 1 or 2 examining facilities, much like that used by the Ranch Hand study* The physical examination will include a standard, good quality review of systems. Multiple laboratories may be used for the various laboratory tests, but each particular test will be performed in a single laboratory. Special emphasis will be given to the clinical and laboratory outcomes which will be chosen during protocol development from among those which are listed later in this outline* �Page 57 Vietnam Experience Study Tentative Timetable This tentative timetable is divided into 2 phases - protocol development and study implementation. However, some tasks which are formally a part of the implementation phase are scheduled to begin during the development phase. This approach is proposed so that there will be no unnecessary delays in the event that the protocol review goes smoothly and according to schedule. Month number 1 for each study phase begins at the time resources are made available to CDC by the VA. Study Phase Month Number Major Milestones Protocol Development recruit new personnel and short-term consultants for protocol development 3 o complete development of protocol 4 o complete peer review of protocol o complete preliminary work with military files for sample selection begin developmental work for contracts for questionnaire administration, clinical and laboratory work complete OMB review complete selection of pilot study samples Study Implementation begin selection of main study samples begin final formatting of questionnaires and clinical instruments begin data collection for main study mortality analysis award contract for questionnaire administration �Page 58 Vietnam Experience Study Tentative Timetable (continued) Study Phase Month Number Major Milestones 7 0 begin questionnaire pilot study 10 0 award contract for clinical and laboratory studies U o begin clinical and laboratory pilot study o evaluate questionnaire pilot study 12 o begin questionnaire main study 16 o evaluate clinical and laboratory pilot study 17 0 begin clinical and laboratory main study 23 o complete study sample selection 32 o complete mortality study data collection 35 o REPORT mortality study analysis 36 o complete questionnaire data collection 41 o complete clinical and laboratory data collection 42 o REPORT questionnaire analysis 47 o REPORT clinical and laboratory data collection �Page 59 Agent Orange Study Tentative Timetable Timetable for this study will parallel the Vietnam experience study timetable in the early phases (i.e., protocol development and review). Because of the extra time required to review military records for determination of Agent Orange exposure, data collection for the 3 study phases (mortality, questionnaire, clinical) will begin approximately 6 months after the comparable phase of the Vietnam experience study. Accordingly, the reports will appear 6 months later: Study Phas6 Study Implementation Month Number Major Milestones 41 o REPORT mortality study analysis 48 o REPORT questionnaire analysis 53 o REPORT clinical and laboratory data collection Tentative List of Items for Health Questionnaire, Physical Examination and Laboratory Analysis The questionnaire and physical examination instruments will be drawn up during the protocol development phase. The following is a list of important elements which will serve as the starting point for development of the final instruments. Questionnaire Information: 1. Locator and Tracing Information 2. Demographic Information 3. Other Potential Confounders: Military History: Drafted vs enlisted status Military occupational specialty Combat vs noncombat experience: Duties, places, dates (develop combat index from casualty rates, # enemy attacks, etc., from sample of records as well as asking men) Area of service Discharge status Tobacco (types of use, amount of use, dates of use) Alcohol (types of use, amount of use, dates of use) Medications (amount of use, dates of use): �Page 60 3. (Continued) Antiraalarials—primaquine, chloroquine, fansidar, dapsone, etc. Antifungals—griseofulvin, etc. Other medications (also include reason for use) Illicit drug use (amount of use, dates of use): Marijuana, barbiturates, amphetamines, opiates, cocaine, PGP, hallucinogens Specific chemical exposures (how, how much, and when exposed; CF.): Agent Orange—include 2,4~D and 2,4,5-T Other herbicides Pesticides, insect repellants Riot control agents Occupational history (type of job, dates, chemical exposures, if any) Hobbies (e.g., chemical exposures, risk-taking behaviors) Habits: L. Breslow's healthy habits, index of social linkage 4. Medical history: Family history: Immediate family: age now or at death; if dead, cause of death; Illnesses requiring hospitalization, surgery, or medication Personal history (before, during, and after military service): Personal physician: name, address, telephone number Specific illnesses (who, what specifically, when, how severe, source of verification): high blood pressure, heart disease, cancer, stroke, lung disease, diabetes, mental or nervous diseases, liver disease, arthritis, repeated infections, malaria, parasitic diseases Hospitalizations (reason, year, duration, source for verification) Surgical procedures (reason, year, duration, source for verification) Blood transfusions (reason, year, source for verification) Injuries (year, severity, source for verification) Allergies (year, severity, source for verification): asthma, rash, hay fever, medication reactions Time lost from work _1 week (reason, year, duration, source for verification) Review of systems: (date, duration, severity when positive response) Weight on discharge from military, 1 year ago, and today General: change in weight (if loss, intentional or unintentional), loss of appetite, weakness Head: headaches, change in hair pattern Eyes: change in vision, irritated eyes Ears: change in hearing, ear noises, ear infections Nose: sinus infections, nosebleeds Mouth: sore tongue, sore throat Neck: swollen glands, goiter (large thyroid), stiffness, pain Chest: shortness of breath, cough, wheezing, phlegm, chest pain, heart attack, heart failure, heart murmur, palpitations Abdomen: difficulty swallowing, vomiting, gallstones, difficulties with digestion, change in bowel habits, blood in bowel movement, hemorrhoids, hernia �Page 61 4. (Continued) Genitourinary: venereal diseases, kidney stones, kidney infections, blood in urine, impotence, decreased sex drive, infertility, children with birth defects Limbs: swelling, change in skin color, joint pain, difficulty with movement, difficulty with coordination, numbness, tingling, pains Neuropsychiatric: concussion, forgetfulness, sleep disorders, paralysis, seizures, dizziness, depression Skin: rashes, boils, acne, scars, sunburns easily, bruises easily 5. Physical examination (CF., NCHS and Ranch Hand physical exam sheets): General: appearance, weight, height, blood pressure, pulse, respiratory rate Head: movements, hair pattern Eyes: movements, conjunctivitis Ears: hearing, infections Nose: polyps, sinusitis Mouth: teeth, tonsils, tongue, cheeks, throat Neck: movement; thyroid enlargement, nodules, tenderness; parotid enlargement or tenderness; cervical lymphadenopathy Chest: movements, bony abnormalities, axillary lymphadenopathy Lungs: rales, rhonchi, wheezes, dullness, hyperresonance Heart: extra sounds, murmurs, rubs, size Abdomen: liver size, spleen size, tenderness (location), masses, hernia, testicular masses, inguinal lymphadenopathy, rectal exam, Back: scoliosis, kyphosis, tenderness (location) Limbs: movements, edema, arthritis, varicose veins, nail clubbing, peripheral pulses The following exams should be done by a dermatologist and a neurologist, respectively: Skin: rash, scars, ulcers, acne, masses, spider angiomata, etc.; Neurological exam: Mental status: Emotional responses: Cranial nerves: Motor systems: gait, movement, tremors, muscle bulk, muscle tenderness Reflexes: Sensory tests: 6. Psychological testing (CF., Ranch Hand set of tests—need consultation): Minnesota Multiphasic Personality Inventory Wechsler Adult Intelligence Scale Reading Subtest of Wide Range Achievement test Halstead-Reitan Neuropsychological Test Batteries Wechsler Memory Scale Cornell Index �Page 62 7. Laboratory tests: Blood: Complete blood count: hematocrit, hemoglobin, red cell count, white cell count and differential, platelet count Liver function tests: SGPT, GGTP, total protein, albumen (SGOT, bilirubin, and alkaline phosphatase not necessary but may occur on SMA-12) Kidney function tests: BUN, creatinine Lipid function tests: total and HDL cholesterol, fasting triglycerides Hepatitis B surface and core antigens Immunoglobulin quantitation: IGG, IGM, IGA, IGE, IGD Two hour post-prandial blood glucose VDRL Free T4 and T3 uptake Serum stored for serological testinG (CF., Ranch Hand positives, melioidosis) Urine: Urinalysis: microscopic and dipstick (protein, glucose, hemoglobin) Urine total porphyrins and porphyrin profile Stool: Qualitative test for blood (during physical exam) Other tests depending on results from Ranch Hand study: Chest X-ray Elee trocard iogram B- and T-lymphocyte quantitation �Page 63 APPENDIX B Literature Review 1. Health Effects of Herbicides and Dioxin 1.1. Dermatologic Effects Chloracne is a refractory skin disease characterized by inclusion cysts, comedones, and pustules, with eventual scarring of the skin, produced by environmental exposure to certain halogenated aromatic compounds in humans (Taylor, 1979). A similar condition is also seen in animals. TCDD is an active skin irritant and produces local lesions resembling human chloracne in the skin of rabbit ears (Kimmig and Schulz, 1957). An analogous hyperkeratosis and modulation of sebaceous structures to keratin cysts was observed in monkeys and hairless mice. Since in these species the skin areas affected by TCDO all lack major hair growth, and, in men, lesions usually do not occur in the follicles of beard hair, it has been suggested that the hair shafts on the unaffected portions of the body may facilitate drainage of sebum and keratinaceous debris (Greig, 1979). After acute exposure to TCDD, blepharitis, loss of fingernails and eyelashes, and facial alopecia were observed in monkeys (McConnell et al., 1978a). Horses accidentally exposed to salvage oil containing TCDD in Missouri had hyperkeratotic skin lesions and hair loss, and dogs, cats, and mice similarly exposed had ulcerative dermatitis and hair loss (Case and Coffman, 1973; Carter et al., 1975). In humans, chloracne is the most frequent and consistent acute health outcome of exposure to TCDD. It is often observed in exposed individuals who have no other apparent health effects. However, since it is usual that only patients with chloracne are studied further, it is not possible to accurately estimate the relative frequency of other adverse effects of exposure. There are, however, reports of individuals without chloracne who developed other acute symptoms possibly related to TCDD exposure (Jirasek et al., 1973; Oliver, 1975). Cases of chloracne were reported after the explosions which occurred at factories in Nitro, West Virginia, in 1949 (Suskind, 1978), in Ludwigshafen, West Germany, in 1953 (Goldmann, 1972, 1973), in the Netherlands in 1963 (Dalderup, 1974; Hay, 1976), in Grenoble, France, in 1966 (Dugois et al., 1968), and in the United Kingdom in 1968 (May, 1973). Chloracne has also been reported in occupational exposures that did not involve explosions. These were reported from factories in Middle Rhein, West Germany (Bauer et al., 1961), Hamburg, West Germany (Kimmig and Schulz, 1957; Schulz, 1957), Grenoble, France (Dugois et al., 1958), Newark, New Jersey (Bleiberg et al., 1964), the U.S.S.R. (Telegina and Bikbulatova, 1970), and Czechoslovakia (Jirasek et al,, 1973). In addition to these industrial exposures, chloracne developed in two government scientists involved in the experimental preparation of TCDD (Oliver, 1975). In 1976, the explosion at the ICMESA factory near Seveso, Italy, resulted in the contamination of a large, densely populated area; 187 cases of chloracne have been reported, mostly in children (Malizia et al., 1979). A few of the individuals exposed to the TCDD-contaminated horse arenas in Missouri may have had chloracne (Carter et al., 1975; Kimbrough et al., 1977). �Page 64 Chloracne may persist for many years. For example, 14 of 122 persons with chloracne following the Nitre accident had lesions evident 28 years later (Crow, 1980). One case remained 18 years after the explosion in Ludwigshafen (Goldmann, 1972). Thirteen years after the explosion in Amsterdam, 10 of 50 original cases remained (Hay, 1976). Of 41 employees surveyed 10 years after the U.K. accident, 22 still had mild chloracne (May, 1982). A followup of 55 subjects with chloracne who had worked in the Czech factory revealed that 15% still had florid manifestations after 10 years (Pazderova-Vejlupkova et al,, 1981). Hyperpigmentation and hirsutism may accompany chloracne. Many of the Newark workers with chloracne also developed hyperpigmentation of the sun-exposed areas of the head, neck, and hands or hirsutism, which was always located on the temples. The severity of these conditions paralleled that of chloracne (Bleiberg et al., 1964; Poland et al., 1971). About one-quarter of the Czech workers with chloracne had either hyperpigmentation or hirsutism of the face or both (Jirasek et al., 1973). Mucous membrane irritation has also been reported in several groups of workers (Schulz, 1957; Poland et al., 1971; Goldmann, 1972). 1.2. Hepatic Effects Hepatic porphyria, a disorder of heme pigment metabolism, can either be inherited or acquired by exposure, in both experimental animals and humans to certain polyhalogenated aromatic compounds, medications, and other environmental factors such as excessive alcohol consumption (Strik, 1979; Kimbrough, 1980). All of these chemicals inhibit uroporphyrinogen decarboxylase in the liver, but not in red blood cells. Porphyria cutanea tarda (PCT) is the most severe form of this type of porphyria. A diagnostic indicator of PCT is the simultaneous increase of both uro- and heptacarboxylic porphyrin in urine. It has been found that chronic hepatic porphyria without clinical symptoms begins with accumulation of these porphyrins in the liver, followed by their gradually increasing excretion in the urine. In PCT, skin findings are often associated with increased porphyrin excretion and include excessive skin fragility, vesiculobullous lesions on sun-exposed areas, hirsutism, and hyperpigmentation. However, it appears that PCT and chloracne are independent syndromes (Poland et al., 1971). Porphyria was observed after exposure to TCDD in rats, mice, and chick embryo cells (Goldstein et al., 1973; Kociba et al., 1976; Sinclair and Granick, 1974). It has also developed in several groups of exposed workers. Eleven of 29 Newark workers with chloracne had abnormal excretion of urinary uroporphyrins; of these, three had definite cases of PCT (Bleiberg et al., 1964). A re-examination of the same plant 6 years later revealed no clinical PCT and only one employee with mild persistent uroporphyrinuria (Poland et al., 1971). At least 11 cases of PCT were reported among Czech workers (Jirasek et al., 1973, 1974). Other hepatic effects of TCDD include structural alterations, changes in serum enzyme levels, and changes in the biliary system, in a number of animal species (IARC, 1977; VA, 1981). Many of the reports of human exposures also mention hepatic effects (see also section on carcinogenicity, below). Liver damage was reported in workers in the factories in Hamburg, West Germany, Grenoble, France, Czechoslovakia, and the U.S.S.R. (Kiramig and Schulz, 1957; Dugois et al., 1958; Jirasek et al., 1974; Telegina and Bikbulatova, 1970). Three workers in Middle Rhein, West Germany, had morphological changes in �Page 65 liver biopsies taken 5 years after their exposure ended (Bauer et al., 1961). Liver enlargement and tenderness were reported after the Nitro explosion, and liver damage and hepatitis were reported after the explosion in Ludwigshafen (Zack and Suskind, 1980; Goldmann, 1972). Hepatomegaly was reported among residents of the contaminated region of Seveso (Pocchiari et al., 1979). Effects on enzyme levels have also been reported in humans. TCDD is known to be a potent inducer of a number of hepatic microsomal enzymes (Huff et al., 1980). Increased levels of urinary d-glucaric acid, an indirect measure of hepatic microsomal enzyme activity, were found in children living in the Seveso area (ideo et al., 1982). Altered levels of other enzymes, mainly transaminases and gamma-glutamyl transferases, were also noted (Pocchiari et al., 1979). A slight elevation in the levels of urinary d-glucaric acid and gamma-glutamyl transpeptidase were also observed in a 10-year survey of U.K. workers (May, 1982). Slightly increased elimination of delta-amino levulinic acid has also been reported (Jirasek et al., 1974; Poland et al., 1971). 1.3. Neurological/Psychological Effects Neurological effects of exposure to 2,4-D have been observed in both experimental animals and man. Myotonia of skeletal muscles was produced by 2,4-D administration to rats, guinea pigs, dogs, and rabbits (Danon et al., 1978; Eberstein and Goodgold, 1979; Drill and Hiratzka, 1953; Hill and Carlisle, 1947). Symptoms of asthenia, lethargy, and ataxia were observed in pigs, calves, rats, and mice (Hill and Carlisle, 1947; Bjorklund and Erne, 1966). Irregularities of EEG pattern have been observed in rats, cats, and dogs as well as demyelinization of the spinal cord (Desi et al., 1962). In humans a number of case reports have described symptoms of peripheral neuropathy following poisoning by 2,4-D herbicides. Typical symptoms observed included asthenia, hypesthesia, and myotonia in the muscles of the extremities, hyporeflexia, and general muscular weakness leading to ataxia. Decreased nerve conduction velocities were measured in some cases (Goldstein et al., 1959; Berkley and Magee, 1963; Wallis et al., 1970; and see VA literature review). Irregularities in EEG patterns were observed in farmers exposed to 2,4-D (Kontek et al., 1973). In a survey of 292 workers in a factory that produced 2,4-D, reports of weakness, fatigue, and headaches were very common (Bashirov, 1969). Neuropsychological effects were reported after most of the human exposures to TCDD. Typical complaints among factory workers included fatigue, headaches, weakness and pain, especially in the extremities, sexual dysfunction, loss of appetite, and irritability (Jirasek et al., 1973; Poland et al., 1971; Baader and Bauer, 1951; Goldmann, 1972; Bauer et al., 1961; Kimmig and Schulz, 1957; Crow, 1980; Dugois et al., 1958; Telegina and Bikbulatova, 1970). Two to three years following their exposure to TCDD, two laboratory scientists had similar complaints, including loss of energy and drive, irritability, visual problems, and diminished sense of taste (Oliver, 1975). Headaches were reported among people exposed to the contaminated horse arenas in Missouri (Carter et al., 1975; Kimbrough et al., 1977). Decreased auditory acuity and decreased sense of proprioception were noted among Newark workers. The Minnesota Multiphasic Personality Inventory (MMPI) was administered to the Newark workers. A significant positive correlation was �Page 66 observed between the severity of active acne and the score on the hypomania scale of the MMPI (Poland et al., 1971). Abnormal EEC patterns were noted among workers in Czechoslovakia and Middle Rhein, West Germany (jirasek et al., 1974; Bauer et al., 1961). Neurological studies were conducted following the Seveso accident. A higher percentage of cases of idiopathic clinical or subclinical neuronal damage was found in the most highly contaminated zone than in zones with lower levels of contamination, for both adults and children. The most frequent pathological signs were detected in the peripheral nervous system. Signs of subclinical neuronal damage included reduced nerve conduction velocity (Boeri et al., 1978; Pocchiari et al., 1979). Altered nerve conduction velocity was more prevalent among exposed individuals with chloracne or increased levels of serum hepatic enzymes than among exposed individuals without these manifestations (Filippini et al., 1981). Of about 200 workers from the ICMESA plant and another factory in the same area who were examined for neurological function, 8 were diagnosed as having polyneuropathy of peripheral nerve fibers (Pocchiari et al., 1979). An increased prevalence of slowed nerve conduction velocities was observed among workers employed in the manufacture of 2,4,5-T and 2,4-D in Arkansas (Singer et al., 1982). 1.4. Immuno1ogica1 Effects Acute and subacute doses of TCDD have produced atrophy of the thymus and other lymphoid tissues with loss of lymphocytes in monkeys, rats, mice, and guinea pigs (McConnell et al., 1978a & b; Vos and Moore, 1974). Changes in thymic weight appeared to be a very sensitive indicator of exposure to TCDD, since decreases in thymic weight occurred at doses which had no effect on body weight in rats, mice, and guinea pigs (Harris et al., 1973). Horses exposed to TCDD-contaminated salvage oil were found to have spleens reduced to one-third the normal size and small and inactive lymph nodes (Case and Coffman, 1973). TCDD has also been shown to suppress immune function in animals, primarily thymic-dependent immune function. Suppression of mitogen responsiveness, skin-graft rejection, and delayed hypersensitivity responses have been observed (Vos and Moore, 1974; Vos et al., 1973; Faith and Moore, 1977). Suppression of these T-cell-dependent immune functions appears to occur without helper cell function being affected; thus, different functional subsets of T-cells seem to be selectively affected (Faith et al., 1978). Sensitivity to the immunosuppressive effect of TCDD appears to decrease with age. Exposure of the developing immune system during pre-, and/or post-natal life results in more severe effects than exposure during adult life (Vos and Moore, 1974; Luster et al., 1979). A slight suppression in humoral immunity has been noted (Vos et al., 1973). Low doses of TCDD, which did not elicit clinical or pathological effects, did reduce host defenses in mice to Salmonella infection, while defense to pseudorabies virus was not affected (Thigpen et al., 1975). Susceptibility to Salmonella was found to result from increased sensitivity to bacterial endotoxin (Vos et al., 1978). Non-specific killing by macrophages or specific killing of Listeria was not impaired by TCDD treatment (Mantovani et al., 1979; Vos et al., 1978). �Page 67 Reports of iramunologic effects following human exposure to TCDD have been very rare. An increased susceptibility to infection was noted among workers following the Ludwigshafen accident (Goldmann, 1972). Following the explosion in Seveso, there did not appear to be an increase in number or severity of childhood infections, nor were results of iramunological tests found to be abnormal (Reggiani, 1979, 1980; Malizia et al., 1979; Pocchiari et al., 1979). 1.5. Carcinogenic Effects Several studies indicate that TCDD is carcinogenic in rodents, producing increased incidence of hepatocellular carcinomas and neoplasms in the lung, hard palate, nasal turbinates, and thyroid of the rat (Kociba et al., 1978; Toth et al., 1979; National Toxicology Program, 1982). Hepatocellular tumors, thyroid tumors, and fibrosarcoma of integumentary tissue have been produced in mice (National Toxicology Program, 1982a & b). TCDD may act as a promoter of liver tumors in the rat (Pitot et al., 1980). An association between phenoxyherbicide exposure in forestry workers and soft tissue sarcoma has been noted in two Swedish case control studies as well as in the combined analysis of four American cohorts of workers industrially exposed to phenoxyherbicides (Coggon and Acheson, 1982; Editorial, 1981). Hardell and Sandstrom (1979) found a significant excess of malignant mesenchymal tumors in individuals occupationally exposed to phenoxyherbicide 10-20 years beforehand (relative risk 5.3, with 95% confidence limits 2.4-11.5). Eriksson et al. (1981) also found a significant association between exposure to phenoxyherbicides and soft tissue sarcoma (relative risk 6.8 with 95% confidence limits 2.6-17.3). The histologic distribution of tumor types in the exposed and unexposed groups was not recorded in either study. Honchar and Halperin (1981) combined individuals from 4 cohorts industrially exposed to phenoxyherbicides and related compounds and found that 3 of 105 deaths had been due to soft tissue sarcoma compared with 0.07% of deaths in the total U.S. white male population aged 20-84. A fourth (recently deceased) case was subsequently reported in one of these cohorts (Cook, 1981). Additionally, three other individuals with soft tissue sarcomas were reported to have worked in 2,4,5-T production facilities (Moses and Selikoff, 1981; Johnson et al., 1981). Other studies of workers exposed to phenoxyherbicides during their application have so far failed to confirm this association (e.g., Coggon and Acheson, 1982). However, in most cases the design of these investigations was such that only very high relative risks for soft tissue sarcoma were likely to be detected. Hardell et al. (1981) found a significant excess of lymphomas in Swedish individuals occupationally exposed to phenoxyherbicides (relative risk 6.0, 95% confidence limits 3.7-9.7). The excess risk was similar for Hodgkin's and non-Hodgkin's lymphomas when analyzed separately. No other epidemiologic studies of this association have been reported. Compromised immunity is the strongest risk factor for development of lymphomas (Greene, 1982). Dioxins have immunosuppressant properties in animal species (see above), which presents an attractive hypothesis for the etiology of their postulated association with both soft tissue sarcoma and lymphomas. �Page 68 At least two epidemiologic studies suggest a slight excess risk of stomach cancers in cohorts exposed to phenoxyherbicides and related compounds. Theiss et al. (1982) reported a significant excess of stomach cancers (3 observed vs. 0.6 expected) in 74 German workers who were exposed to trichlorophenol and dioxin 20 years before. Axelson et al. (1980) observed an apparent excess of stomach cancer (3 observed and 0.71 expected) among 348 railroad workers exposed to phenoxyherbicides and amitrol. Hardell et al. (1982) reported that exposure to phenoxy acid herbicides doubled the risk of nasal and nasopharyngeal cancer (relative risk 2.1, not statistically significant). The controls used for this study were the same as those used in the previously mentioned Swedish studies of sarcomas and lymphomas. Tung reported that primary liver cancer occurred in excess in Vietnam as a result of Agent Orange exposure of the general population, but this reported excess was not verified when his report and pathologic specimens were reviewed (VA lit rev., 1981). Even though human liver damage has been reported as a result of dioxin exposure (see above), no excess liver cancer has been reported. 1.6. Reproductive Effects The reproductive effects of 2,4-D, 2,4,5-T, and TCDD, alone or in combination, have been examined in a number of different animal species. The effects are variable, depending on dosage, species, and strain. Only animal studies of the effects of 2,4,5-T with levels of TCDD contamination which either are unknown or known to be at least 1 ppm and of the effects of combinations of 2,4-D, 2,4,5-T, and TCDD will be discussed, in the light of the composition of Agent Orange. A study of the effect of exposure of male mice to contaminated 2,4,5-T before mating with unexposed females showed no effect on the loss of fetuses before or after implantation (Buselmaier et al., 1972). Lamb et al. (1980) examined the effects of "simulated Agent Orange" — i.e., mixtures of 2,4-D, 2,4,5-T, and TCDD — administered to male mice followed by mating to untreated females. No effects were reported in fertility, implantation, fetal malformations, germ cell toxicity, sperm concentration, motility, or abnormalities and survival of offspring. Most of the reproductive studies in animals have involved exposure only of the female after conception. In monkeys, fetal size was reduced but no malformations were observed (Wilson, 1971). In the rat, low doses of 2,4,5-T produced cystic kidney and intestinal hemorrhage (Courtney et al., 1970; Sparschu et al., 1971). A slightly increased incidence of cleft palate in the rat was reported in one study (VA, 1981 lit. rev.). 2,4,5-T administered throughout gestation produced maternal toxicity, fetal death or decreased fetal growth (Hall, 1972). In the mouse, 2,4,5-T produced cleft palate, and cystic kidney, the necessary dosage depending on the strain (Bionetics, 1968; Courtney et al., 1970; Gaines et al., 1974). In the hamster, cleft palate was rarely encountered; instead abnormal cranial development was observed (Collins et al., 1971). �Page 69 Reproductive outcomes have been examined after many human exposures. However, the significance of most of these studies is questionable because of limitations in study design, population size, and inadequate handling of confounding factors. Pazderova-Vejlupkova et al. (1980) considered the frequency of abortion to be normal among wives of workers in the Czech factory. Following the explosion at Seveso, no increase in congenital malformations or developmental abnormalities was noted, but it was not possible to assess the frequency of spontaneous abortions due to an increase in elective abortions following the accident, and no baseline data were available for miscarriages (Reggiani, 1979; Homberger et al., in VA lit. rev.). In the U.S.A., a study of the incidence of spontaneous abortions among women whose husbands were occupationally exposed to 2,4-D as farmers, forest workers, or herbicide applicators revealed no overall association (SRI International, 1981). Human miscarriages near a spray project near Globe, Arizona, were found not to be related to herbicide use; a similar lack of association was found with human malformations in Swedish Lapland (Binns and Balls, 1971; Advisory Committee, 1971). In Arkansas, facial clefts were not associated with the agricultural use of 2,4,5-T (Nelson et al., 1979). A study of birth defects in children born to Long Island Railroad maintenance employees exposed to 2,4,5-T used for weed control revealed that all major birth defects combined and inguinal hernia were less frequent than expected. An excess observed for metatarsus adductus and tear duct obstruction probably resulted from variability in diagnosing these "minor" defects (Honchar, 1982). Reproductive outcomes of wives of Dow Chemical employees exposed to dioxins were surveyed. No statistically significant association between exposure and spontaneous abortions, stillbirths, infant deaths, and congenital malformations was observed (Townsend et al., 1982). The reported association between 2,4,5-T spraying and an increased incidence of miscarriage in the Alsea basin of Oregon (EPA, 1979) has been severely criticized (Wagner et al., 1979; Mantel, 1979). A number of studies of reproductive outcomes were conducted in Australia and New Zealand. A study in Australia revealed no relationship between 2,4,5-T use and birth defects (Aldred et al., 1978). Another showed a correlation between the season of conception of babies with neural tube defects and the season of maximum 2,4,5-T spraying; a correlation was also found between neural tube defects in animals and 2,4,5-T (Field and Kerr, 1979). Two studies in New Zealand found no association between 2,4,5-T exposure and neural tube defects (McQueen et al., 1977; Hanify et al., 1981). One of these also found no association with cleft lip and palate or malformations of the heart or male genitalia, although it did reveal an association with talipes (malformations of the foot). A study in Western Australia that suggested an association between cleft lip and palate and herbicide exposure (Brogan et al., 1980) has been criticized on methodologic grounds (Bower and Stanley, 1980). A survey of ground agricultural sprayers showed no differences in the occurrence of malformations, stillbirths, miscarriages, or ectopic pregnancies (Smith et al., 1981). The reports of human birth defects alleged to result from exposure to Agent Orange, which appeared in South Vietnamese newspapers in 1969, caused public and scientific furor (Advisory Committee, 1971; Young et al., 1978). In response, two independent surveys of South Vietnamese hospital records were conducted. An apparent increase in certain birth defects relative to others, which seemed to be associated with periods of herbicide spraying, was noted by �Page 70 Meselson et al. (1971). Cutting et al. (1970) found no increased incidence of congenital abnormalities, stillbirths, and hydatidiform moles with heavy herbicide spraying. However, the conclusions of both of these studies were seriously limited by incomplete and unrepresentative sampling of births, unreliable birth records, and inadequate estimation of exposure (Advisory Committee, 1971). A subsequent study found an increased prevalence of isolated cleft palate and spina bifida compared with earlier years before widespread defoliant use, which might, however, be attributable to better case-finding and referral (Herbicide Assessment Commission, 1970; Nelson et al., 1979). Tung et al. (1971) and Rose and Rose (1972) reported on malformations and abortions among South Vietnamese refugees in North Vietnam. Lack of specific information about exposure and the lack of an unbiased selection procedure preclude any causal inferences. Studies conducted in South Vietnam in 1972 and 1973 by the National Academy of Sciences (1974) found no conclusive evidence of association between human birth defects and herbicide exposure, although study limitations were recognized. A report has just been released on a large study (Donovan et al., 1983) designed to determine if Australian Vietnam veterans are at increased risk of fathering babies with birth defects. Vietnam veterans had no greater risks than veterans who served elsewhere or than men who were not veterans. 1.7. Other Effects Gastrointestinal problems have been reported after a number of human exposures. A health survey of workers involved in 2,4-D production revealed that about half complained of dyspepsia, abdominal pains, and constipation (Bashirov, 1969). About 30% of the workers studied at the Newark plant complained of gastrointestinal symptoms (nausea, vomiting, diarrhea, abdominal pains, or blood in stool) (Poland et al., 1971). Digestive disorders were reported among workers in the factories in Grenoble, France, and in Hamburg and Middle Rhein, West Germany (Dugois et al., 1958; Schulz, 1957; Bauer et al., 1961). Gastrointestinal symptoms, including abdominal pains and indigestion, were among the delayed symptoms which developed 2 to 3 years after TCDD exposure in two of the three government scientists in England (Oliver, 1975). High levels of serum cholesterol and lipids were also commonly reported among exposed workers. Serum lipids tended to be high among workers following the explosion at the Nitro factory (Suskind, 1978). Ten percent of Newark workers had elevated serum cholesterol levels (Poland et al., 1971). Hyperlipemia and hypercholesterolemia were reported among workers in Grenoble (Dugois et al., 1958). Similar findings were described for the Czech workers, who also exhibited elevated levels of pre-beta lipoprotein and of total blood proteins (Jirasek et al., 1974; Pazderova-Vejlupkova et al., 1980, 1981). All three of the English scientists had hypercholesterolemia (Oliver, 1975). Walker and Martin (1979) reported high cholesterol and triglyceride levels and low high-density-lipoprotein levels in a small group of exposed workers. 2. Diseases Affecting U.S. Troops in Vietnam This section is included to provide background on the health of U.S. servicemen while they were stationed in Vietnam. Fifty-six to seventy-four percent (mean 70.6%) of hospital admissions during the Vietnam war were for �Page 71 medical disorders, as compared with battle casualties (15.6%) and non-battle Injuries (13.8%), during the period 1965-69 (Ognibene and Barrett, 1982). Despite this fact, the average annual disease admission rate (351 per 1,000 per year) was one-third lower than for the China-Burma-India and Southwest Pacific theaters In WWII, and 40% less than for the war In Korea (Neel, 1973). Malaria has been Identified as the most significant medical problem, accounting for the greatest number of man-days lost from duty during the war. The emergence of a chloroqulne-reslstant form of malaria, P. falclparum malaria, led to the use of DapsoneR (4,4 •-diaminodiphenylsulfone), which Is also used to treat leprosy (Neel, 1973). Infectious hepatitis did not pose a major problem during the Vietnam war, as It did In previous wars. The Incidence of hepatitis (6.9 cases per 1,000 per year) varied with the Intensity of combat operations and with troop Interaction with the civilian population (Neel, 1973). In Vietnam, serum hepatitis was of more concern, occurring most commonly among men who received multiple blood transfusions related to battle injury or among those using Illicit drugs Intravenously (Ognibene and Barrett, 1982). Diarrheal disease rates were also lower compared with earlier wars. The prevalence rate ranged from 69 per 1,000 in 1965 to 35 per 1,000 in 1969. Diarrheal diseases may have been related to viruses, bacteria or parasitic agents, but the cause of most cases could not be identified. Troops at greatest risk were those who were unacclimatized and those under combat conditions. Incidence peaked in May or June, corresponding with the monsoon season (Neel, 1973). Skin diseases were quite prevalent among troops in Vietnam. Those cases severe enough to require hospitalization or retention in quarters varied from 30 per 1,000 in 1965 to 20 per 1,000 in 1968. In 1970, however, skin problems Increased again, to 30 per 1,000. The reason for the Increase is unexplained* The three major skin problems identified were superficial fungal infection, bacterial Infection, and immersion foot (Neel, 1973; Allen, 1977). Plague and cholera, endemic in the Vietnam population, did not pose a significant problem for U.S. troops. Melioidosis, an infectious disease of humans and animals endemic in tropical areas, presented a problem to U.S. physicians unfamiliar with its diagnosis or treatment. Two hundred and thirty cases, diagnosed between 1965 and 1971, resulted in 14 deaths (Neel, 1973). The problem of fever of undetermined origin (FUO) presented some of the most challenging diagnostic dilemmas for military physicians in Vietnam. The diagnosis of FUO ranked second only to venereal disease. During the period 1966 through 1969, 58 cases per 1,000 were reported each year, including hospitalized and non-hospitalized patients (Ognibene and Barrett, 1982). Venereal diseases have been prevalent during most military engagements. In Vietnam, it led other common medical problems in prevalence from 1965 to the conclusion of the war. Gonorrhea accounted for 90% of all venereal disease cases* The second most frequently occurring condition of venereal origin was chancroid (Ognibene and Barrett, 1982). Neuropsychiatric diseases did not differ appreciably among troops serving in Vietnam and those serving elsewhere until 1968. During this year, the prevalence of psychosis, psychoneurosis, and of character and behavior �Page 72 disorders increased among all army troops and particularly among those stationed in Vietnam and became the second leading disease problem by 1970. Concomitantly, the problem of drug abuse escalated during this period, especially among younger, lower ranking enlisted men (Neel, 1973). 3. Current Health of Vietnam Veterans Very little is known about the health of Vietnam veterans relative to the health of other men of similar age. Some indication of veterans' and others' perceptions about the veterans' health can be found in the reports of Bogen, 1979; Stellman and Stellman, 1980; Texas Dept. of Health, 1983; UCLA-VA Protocol literature review; and Wolfe, 1980. The most frequently reported conditions include dermatologic disorders, neurologic and psychologic disorders (including numbness and tingling in the extremities, headaches, fatigue, depression, memory loss, sleep disturbances, and sexual dysfunction), reproductive problems (birth defects, miscarriages, abortions, reduced fertility), cancer, gastrointestinal disorders, infections, hypertension, hepatic hematologic, genitourinary, respiratory, and cardiovascular problems. Although there is a lack of data on organic disease outcomes among Vietnam veterans, there are a number of reports on the occurrence of health-related outcomes — outcomes which may be considered by some to be disease outcomes and by others as possible causes or effects of disease. Several large surveys have been conducted which provide psychological and sociological data on Vietnam veterans, veterans who served in the Vietnam era but not in Vietnam, and contemporary non-veterans (Starr et al., 1973; Martindale and Poston, 1979; Hammond, 1980; Harris and Assoc., 1971; Egendorf et al., 1981). These surveys present objective data concerning several aspects of social adjustment, subjective reports of psychological adjustment, and attitudes held by and about Vietnam era veterans* Although these surveys employed a variety of methods and focused on different aspects of adjustment, it can be concluded from this literature that Vietnam veterans have encountered more problems in adjusting to civilian life than the other men (Figley, 1977; 1978). The general areas of observed or suspected sociological differences among Vietnam veterans, other Vietnam era veterans and non-veterans include educational and occupational status, stress-related psychological difficulties, drug and alcohol use, medical problems, and arrests (Boscarino, 1981; Boscarino and Figley, 1981; Segal, 1977; Borus, 1975; Cover and McEaddy, 1974; Stinson, 1979; O'Brien et al., 1980; Mintz et al., 1979). These problems have been found to vary among subgroups of these populations defined by ethnicity, exposure to combat, urban or rural residence, and period of service in Vietnam (Egendorf et al., 1981; Penk et al., 1981). Post Traumatic Stress Disorder (PTSD) and its association with Vietnam service, exposure to combat, and drug and alcohol use has been widely investigated (Roberts et al., 1982; Boraan, 1982; Lipkin et al., 1982; Frye and Stockton, 1982; Wilson & Kruass, 1982; Boscarino, 1980; 1981; Helzer et al., 1979; DeFazio et al., 1975; Horowitz, 1975). PTSD is thought to be a very common condition among Vietnam veterans (Wilson, 1980). However, large-scale psychiatric epidemiology research, which treats PTSD as a distinct diagnosis, has not yet been reported. Reliable estimates of the prevalence of PTSD in �Page 73 the Vietnam veteran population cannot be derived from the current literature because of the frequent use of unusual (e.g., treatment seeking) samples and because symptom frequencies instead of validated diagnostic criteria have been used as outcome measures. 4. Long-Term Health Status of Servicemen and Veterans This literature was reviewed to provide background for the Vietnam Experience study. The writers of these protocols expected to find a rich literature, but did not.* Numerous health studies of veteran populations have been conducted, but there are few, if any, which deal with long-term health effects of the general war experience. Disease incidence and prevalence among army personnel is well documented for World War II (WWII) (Anderson, 1968), the Korean War (Army Medical Service Graduate School, 1954), and the Vietnam conflict (Ognibene and Barrett, 1982) (see part 2, this Appendix); however, these reports cover only the period of military action. *For reports of studies on the long-term health effects of war experience, we reviewed the Cumulated Index Medicus for the years 1975 through March 1983. In addition, several computer-based literature searches were conducted against these on-line data bases: Medline, 1966-83; Cancerlit, 1963-83; American Statistics Index, 1974-82; Social Science Citation Index, 1972-83; Psych Info, 1967-83; and Sociological Abstracts, 1963-83. The holdings of the libraries maintained at the Centers for Disease Control, Veterans Administration (VA) Hospital (Atlanta), VA Central Office (Washington) and Emory University School of Medicine were reviewed for appropriate reports. Finally, relevant studies completed on veteran populations by the Medical Follow-up Agency of the National Research Council within the National Academy of Sciences were included in the literature search. When relevant studies were identified, we used a branching technique to search for other cited references. A total of 135 journal articles and books were brought to CDC offices and reviewed. �Page 74 A summary of the studies reviewed follows, even though they are not especially useful for the task at hand. Hawryzluk (1975) studied prevalence ratios of diagnosed conditions among 813 army officers. Hearing loss, musculoskeletal disorders, and skin disorders were among the most frequently occurring medical problems. This study was limited to officers, most of whom were between 33 and 37 years old and had had 10-14 years of military service. They were selected for leadership positions and for their potential ability to do college work; thus, they were probably not representative of the general military population. Medical records from the Armed Forces and the VA offer opportunities for followup studies. The Armed Forces system records all illnesses and injuries, even minor ones, among its active duty members, and it stores the clinical records in a central repository when the individual is separated from service. In the VA system, records documenting most of the agency's contacts with a veteran are maintained in a single file. Because benefits to veterans are many and varied, the VA maintains contact with most veterans, and many thousands of records are thus accessible for study (DeBakey and Beebe, 1962), (Beebe, 1951), (Cohen, 1953). However, because only a fraction of veterans receive their health care at VA facilities, and because those who do may be less educated and have more severe service-connected physical and mental disabilities, the records are of questionable usefulness for epidemiologic purposes, since their health experiences may not reflect those of the overall veteran population. Armed Forces and VA records have been used for clinical followup studies of various medical and traumatic conditions, such as leprosy (Brubaker et al., 1969), rheumatic fever (Engleman et al., 1954), missiles in the heart (Blano and Beebe, 1966), and psychoneuroses (Brill and Beebe, 1951)• These studies have been conducted for the purpose of describing the natural history and progression of the disease or condition and were conducted without control groups. Other studies with control groups, on the basis of the Armed Forces and VA data bases, have been directed at the veteran population receiving health services through the VA system, for example: studies of amyotrophic lateral sclerosis (Kurtzke and Beebe, 1980), asthma (Robinette and Fraumeni, 1978), scrub typhus (Elsom et al., 1961), coronary heart disease (Hrubec and Zukel, 1974), lumbar disc lesions (Hrubec and Nashold, 1975), splenectomy (Robinette, 1977), infectious mononucleosis (Miller and Beebe, 1973), cirrhosis of the liver (Beebe and Simon, 1970), esophageal cancer (Rogers et al., 1982), traumatic limb amputations (Hrubec and Ryder, 1980), and learning and reaction time (Milligan and Powell, 1981). Generally, the controls for these studies have been other veterans. Since the diseased and control veterans in these studies were not stratified with respect to their combat participation, the effect of that experience on the occurrence of the disease or its clinical course cannot be evaluated. Veterans or their families have been participants in several studies on the effect, on subsequent health, of exposure to certain risk factors. Wallis (1968) reported on stress in service families, but his study did not include control families. Other studies have examined the effect on veterans of exposure to adjuvant influenza virus vaccine (Beebe et al., 1972), microwave radiation (Cleary et al., 1965), mustard gas (Beebe, 1960), (Norman, 1975), and smoking (Rogot and Murray, 1980). These studies included control groups, �Page 75 but they were also selected from among other veterans. For the reasons discussed above, these data cannot be used to evaluate the effect of war service. The literature contains reports from several studies that examined the morbidity and mortality experience of prisoners of war (POW's). Nefzger (1970) found that standardized mortality ratios and death rates indicated a clear early excess of deaths among prisoners held by the Japanese in WWII. Prisoners from the European and Mediterranean theatres of WWII did not have an adverse mortality experience to 1965. Keehn (1980) followed the same groups through 1975 and found that their increased risks of death, though diminished over time, persisted for 9 and 13 years, respectively. Mortality in Korean War prisoners has been more like that in Pacific than European WWII prisoners (Nefzger, 1970). Mortality from tuberculosis and from trauma contributes to the increase among Pacific ex-prisoners, whereas for Korea the increase is limited to trauma. An excess of deaths due to cirrhosis of the liver was apparent in all three former prisoner groups, WWII (Europe, Pacific) and Korean, from about the 10th followup year (Keehn, 1980). Beebe (1975) studied morbidity, disability, and maladjustments among WWII and Korean prisoners and compared them with veteran controls from the same wars who were not taken captive. In this study, sequelae of the POW experience were both somatic and psychiatric and were of greatest extent and severity among Pacific WWII POW's. Among European WWII POW's, only psychiatric sequelae were apparent. Somatic sequelae were most prevalent in the early years after liberation, but for Pacific WWII POW's they persist in the form of higher hospital admission rates for many specific causes. Klonoff et al. (1976) investigated the long-term or residual effects resulting from severe and extended exposure to stress among POW's captured in Japan (high-stress group) or Europe (low-stress group) during WWII. The low-stress group was divided into long-term and short-term internment periods. Neuropsychological, psychiatric, and physical/neurological outcomes were compared, and significant differences were found among these three groups. The high-stress group scored significantly lower in operational intelligence, exhibited more signs of psychiatric maladjustment, and had more physical illnesses, especially of the neurological and musculoskeletal systems. Residual effects increased in proportion to length of internment, though numbers in each category were small when stratified in this way. The authors concluded that terms such as "survival syndrome" (Chodoff, 1963) and "war neurosis" (Maskin, 1966) describe identifiable phenomena with long-term residual effects (Klonoff et al., 1976). Davies (1978) found an excess of leukemias, lymphomas, myelomas, and polycythemia vera among Australian servicemen with overseas and tropical area service as compared with those serving in temperate Australia; however, he did not control for confounding variables (such as age) and, for some controls, the area of service was doubtful. A diagnosis of malaria and/or an interaction of nitrates and nitrites with the malaria prophylactic drug chloroquine were suggested as possible risk factors. In a followup study, Giles et al. (1980) investigated the possibility that exposure to malaria may have led to later development of lymphoma in 62 men resident in Tasmania, Australia, and found no association. �Page 76 In two studies which covered 29 years (1946-1974), Jablon and Miller (1970, 1978) found no statistically significant differences between army x-ray technologists (n=6,560) and controls (n=6,826) who served as medical, laboratory, or pharmacy technologists for total deaths from cancer, individual site of cancer, or deaths from other causes. Norman et al. (1981) investigated exposure to tetrachloroethane by comparing age-specific mortality among 1,099 males assigned to chemical processing companies during WWII and 1,319 veterans not involved in the impregnation process of protecting clothing against mustard gas. Overall cancer mortality for exposed subjects was 1.26 times higher than for controls. The risks for leukemia, lymphoma, and cancers of the genital organs were moderately elevated, but the numbers were small and no significant excesses were observed. The Medical Followup Agency of the National Academy of Sciences National Research Council established a Twin Registry comprising 16,000 pairs of white male twins, both members of which had been in military service, mainly in WWII. This data base has provided information for the study of multiple sclerosis (Bobowick et al., 1978), cardiovascular and respiratory symptoms (Cederlof et al., 1969), (Hrubec et al., 1973), psychopathology (Pollin et al., 1969), (Allen and Pollin, 1970), (Hoffer and Pollin, 1970), (Stabenau et al., 1970), intraocular pressure (Schwartz et al., 1972, 1973), corticosteroid response (Schwartz et al., 1973), allergy (Bazaral et al., 1974), skin diseases (Lynfield, 1974), hypertension (Oglesby, 1975), headache (Ziegler et al., 1975), plasma cholesterol and triglycerides (Christian et al., 1976), personality traits (Horn et al., 1976), earnings (Taubman, 1976), dietary intake (Fabsitz et al., 1978), weight changes (Fabsitz et al., 1980), electrocardiographic characteristics (Havlik et al., 1980), alcoholism (Hrubec and Omen, 1980), and familial factors in early deaths (Hrubec and Neel, 1981). These studies have not classified the veterans according to their combat experience. Seltzer and Jablon (1974) found evidence for a "healthy warrior" effect when they examined the effect of health selection at induction on subsequent cause-specific mortality in a series of 85,491 white male WWII U.S. Army veterans followed for 23 years, 1947-1969. They found that mortality rates were well below those of the general population during the first few years after discharge. After 23 years the mortality rates of the veterans were still lower than, but approaching, those of the general population. The effect of military selection varied considerably according to the nature of the cause of death. Three studies have demonstrated an association between mortality and military rank at separation from military duty. Keehn et al. (1978, 1974) and Seltzer and Jablon (1977) found that mortality during 24 years following separation declined with each successive advance in rank through the enlisted grades. Furthermore, mortality of privates was very close to expectation based on population rates; non-commissioned officers had a 23% advantage and commissioned officers about a 40% advantage. The advantage held for deaths from all causes and also for most specific causes examined. Over the 24-year period of followup, the tendency for the differences to diminish was only small. �Page 77 In summary, many health studies have been conducted on veteran populationst but because of the lack of control groups, the selection of control groups from among veterans who were not classified as to their combat experience, and the selection of study subjects from specific military occupational specialties, the studies are not useful for evaluating the overall effect of war service. CDC's review of this literature revealed little which could be used to generate specific hypotheses about health effects of military service in the Vietnam war. �Page 78 APPENDIX C SAMPLE SELECTION USING TELEPHONE RANDOM DIGIT DIALING Random digit dialing is a telephone sampling method that produces a random sample of households with telephones, regardless of whether or not the number is listed in the telephone directory. It appears to be an efficient and inexpensive means of obtaining an unbiased random sample, and a preferable alternative to time-consuming and costly door-to-door screening and to random selection of numbers from telephone directories or specially compiled lists. The latter approach misses unpublished and new listings and requires the difficult task of removing duplicates when large geographic areas and multiple overlapping directories and lists are involved. Further, since 90.2% of all U.S. households had telephones in 1976 (thought to be around 95% in 1983), biases attributable to underrepresentation of those households that do not have telephones are not likely to affect results appreciably (Klecka and Tuchfarber, 1976). One factor to be aware of, however, is that availability of telephones is related to income. . According to the 1970 Census of Population and Housing, 76% of households with incomes $5,000 had telephones, compared with 95% of households with incomes I$25,000; 89% of white households had telephones, compared with 70% for black households (Waksberg, 1978). Random digit dialing methods range from dialing a 7- or 10-digit random number to compiling a listing of area codes plus 3-digit exchanges used within the geographic bounds from which a study sample is to be drawn and randomly appending the last 4 digits. The 7- and 10-digit random numbers are estimated to produce households for only 1 in 30 and 1 in 200 numbers dialed, respectively (Cooper, 1964; Glasser and Metzger, 1972). Sampling within the listing of area code plus 3-digit exchanges involves one of several approaches to randomly append the last 4 digits and tp deal with non-residential and not-in-service numbers. Klecka and Tuchfarber (1974a) report that the proportion of not-in-service numbers ranged from 37.3% in an urban setting to 70.6% in a rural region for 3 random digit dialing samples; and the proportion of business numbers were 11.3% and 3.2%, respectively. Cooper (1964), who uses blocks of 3-digit exchanges plus 1 digit and randomly selects the remaining 3, reports 32% of the numbers were ineligible. Waksberg (1978) contends that simple random sampling within existing exchanges is inefficient, since about 80% are businesses, institutions, government, or not in service. Waksberg's method seems to eliminate making large numbers of nonproductive calls to non-residential and not-in-service numbers by making multiple calls within a block of numbers (block=area code + exchange + 2 random numbers) only if the first number dialed within that block is residential. To support the hypothesis that random digit dialing yields an unbiased sample, such a sample must be scientifically compared with samples drawn by conventional means in the field. In 1974, Klecka and Tuchfarber (1976) compared their random digit dialing sample on crime victimization of 800 households and 1,685 respondents in Cincinnati, Ohio, with the Census Bureau's survey of 9,708 households and 19,903 respondents. Race, age, sex, education, income, household density of persons over 12 years of age, and ownership status of the residence were among the demographic variables examined. Excepting education, there were no statistically significant differences between the two populations when tested by chi-square. Thus, the authors concluded that random digit dialing and Census Bureau's complex approach had produced samples from the same population. References cited above and others documenting the efficacy of random digit dialing are found in section 12. �Page 79 APPENDIX D TOPICAL LIST OF QUESTIONNAIRE ITEMS* FOR AGENT ORANGE AND VIETNAM EXPERIENCE STUDIES ADMINISTRATIVE Name Identification Numbers Military Service Number Social Security Number Telephone Number Interviewer Name Date of Interview Quality of Interview Names and addresses of friends who will know future whereabouts SOCIODEMOGRAPHIC Date of Birth Place of Birth Current Residence Race/Ethnicity Marital History Education Religion Occupation and Income Problems in Obtaining Employment MEDICAL Height and Weight General Health Status All Hospitalizations and Operations Physician Treatment, Physician Diagnosis, or Self-Diagnosis of: Neurologic Disorders Psychologic Disorders Impaired Fertility Endocrine Diseases Cardiovascular Diseases Cancer Gastrointestinal Disorders Genitourinary Disorders Respiratory Diseases Musculoskeletal Condition Dermatologic Conditions Other Complaints Trauma Reproductive History Blood Transfusions *Some data items listed may be derived from military records. �Page 80 ENVIRONMENTAL AND OCCUPATIONAL EXPOSURES Smoking Alcohol Abbreviated Occupational History Focusing on Exposures to Herbicides Illicit Drug Use MILITARY HISTORY Drafted/Enlisted Countries of Assignment Occupational Duties Combat Intensity Injuries, Wounds in Service Herbicide Exposure �Page 81 APPENDIX E TOPICAL LIST FOR EXAMINATION AND LABORATORY TESTING* AGENT ORANGE AND VIETNAM EXPERIENCE STUDIES PHYSICAL EXAMINATION The physical examination will be modified from those of the National Center for Health Statistics' Health and Nutrition Examination Survey and the Ranch Hand Study, with special attention given to the dermatologic and neurologic systems. General: habitus, weight, height, blood pressure, pulse, respiratory rate Skin: rash, scars, ulcers, acne, masses, spider angiomata, pigmentation Head: movements, hair pattern Eyes: movements, fundi, Snellen testing of acuity, conjunctiva, icterus Ears: audiometry, otoscopic exam Nose: polyps, sinusitis Mouth: teeth, tonsils, tongue, cheeks, throat, gingiva Neck: thyroid and parotid palpation, cervical lymphadenopathy Chest: movements, bony abnormalities, axillary lymphadenopathy Lungs: rales, rhonchi, wheezes, dullness, hyperresonance Heart: extra sounds, murmurs, rubs, size Abdomen: liver and spleen size, tenderness, masses, hernias, testicular size and masses, inguinal lymphadenopathy, rectal exam Back: scoliosis, kyphosis, tenderness Limbs: movements, edema, arthritis, varicosities, nail clubbing, peripheral pulses, lymph nodes Neurologic: mental status, cranial nerves, motor system, reflexes, sensory deficits, nerve conduction studies (conduction evaluation only for Agent Orange study) PSYCHOLOGIC AND NEUROPSYCHOLOGIC TESTING Minnesota Multiphasic Personality Inventory Diagnostic Inventory Schedule Psychiatric Epidemiology Research Interview Battery from Halstead-Reitan Neuropsychological Tests Armed Forces Qualification Test—this is the intelligence test given to the veterans on their induction into service Wechsler Memory Scale * May be modified as a result of consultations to take place in late 1983 and early 1984 with experts in several specialties, e.g., neurology, immunology, psychology. �Page 82 LABORATORY TESTING BLOOD: Complete Blood Count: hematocrit, red cell count, white cell count and differential, platelet count Fasting Blood Glucose Cholesterol and Triglycerides Greatinine Bilirubin and GGPT Thyroxine Hepatitis B Core Antibody Serum Stored for Future Serologic Testing URINE: Protein Glucose Hemoglobin Porphyrins STOOL: Qualitative Test for Occult Blood MISCELLANEOUS: Delayed Cutaneous Hypersensitivity Battery: Mumps Candida Tuberculin Streptococcus Proteus Diphtheria Tetanus Control �Page 83 APPENDIX F TOPICAL LIST OF QUESTIONNAIRE ITEMS FOR SELECTED CANCERS CASE-CONTROL STUDY ADMINISTRATIVE Name Identification Numbers Military Service Number Social Security Number Telephone Number Interviewer Name Date of Interview Quality of Interview Friends who will know future whereabouts SOCIODEMOGRAPHIC Date of Birth Place of Birth Current Residence Race/Ethnicity Marital Status Education Religion Occupation and Income FAMILY HISTORY OF CANCER Occurrence of soft tissue sarcomas, lymphomas, and other cancers in first-degree (parents, siblings, and children) and second-degree (aunts, uncles, and grandparents) blood relatives and spouses. �MEDICAL Page 84 Height and Weight Possibly Predisposing Conditions Immune Deficiency Diseases Rheumatoid Arthritis Other Cancers Celiac Disease/Gluten Enteropathy Hemophilia Infectious Mononucleosis Neurofibromatosis Trauma Medical Exposures Immunosuppresive Therapy X-irradiation Dilantin Iron Dextran Blood Transfusions Surgery, Hospitalizations, Long-term Medications Medical Care Utilization ENVIRONMENTAL AND OCCUPATIONAL EXPOSURES Smoking Alcohol Lifetime Occupational History, Including Probes to Exposures Such As: Asbestos Herbicides Pesticides Irradiation Organic Solvents Vinyl Chloride Benzene Arsenicals Wood dust Illicit Drug Use MILITARY HISTORY Drafted/Enlisted Training Countries of Assignment Military Occupational Specialty Occupational Duties Combat Intensity Herbicide Exposure Use of trade names is for identification only and does not imply endorsement of the Public Health Service or the Department of Health and Human Services. �Page 85 12. References Advisory Committee on 2,4,5-T. Report to the administrator of the Environmental Protection Agency, Washington, D.C., May 7, 1971. Aldred JE, Belcher RS, Christophers AJ, Clements A, Danks DM, et al. Report of the consultative council on congenital abnormalities in the Yarram district. Presented to both houses of the Australian Parliament. 1978. Allen AM. Internal Medicine in Vietnam, Vol. I. Skin diseases in Vietnam, 1965-72. United States Army. Washington, B.C., 1977. Allen MG, Pollin W. Schizophrenia in twins and the diffuse ego boundary hypothesis. Am J Psychiatry 1970;127:437-42. Anderson RS (edit). Internal medicine in World War II, Vol. III. Infectious diseases and general medicine. United States Army. Washington, D.C., 1968. Anderson TW. An Introduction to Multivariate Statistical Analysis. Wiley & Sons (Eds.), 1958, New York. John Army Medical Service Graduate School. Recent advances in medicine and surgery: Japan and Korea (presentations). Vol. II. Washington, D.C. 1954. Axelson 0, Sundell L, Anderson K, Edling C, Hogstedt C, Kling H. Herbicide exposure and tumor mortality. Scand J Work Environ Health 1980;6:73-9. Baader EW, Bauer HJ. Industrial toxication due to pentachlorophenol. Surg 1951;20(6):286-90. Ind Med Bashirov AA. Health conditions of workers producing herbicides of amine salt and butyl ester of 2,4-D acid. Vrach Delo 1969;10:92-5. Bauer H, Schulz KH, Spiegelberg U. Berufliche vergiftungen bei der herstellung von chlorphenol-verbindungen. Arch Gewerbepath Gewerbehyg 1961;18:538-55. Bazaral M, Orgel HA, Hamburger RN. Genetics of IgE and allergy: Serum IgE levels in twins. J Allergy Clin Immun 1974;54:288-304. Beebe GW. Follow-up studies of World War II and Korean War prisoners. Am J Epidemiol 1975;101:400-22. Beebe GW. Lung cancer in World War I veterans: Possible relation to mustard-gas injury and 1918 influenza epidemic. JNCI 1960;25:1231-52. Beebe GW. Medical records and the Army, Navy and Veterans Administration follow-up program. (Presentation at Army Medical Service Graduate School.) Feb. 1951. Beebe GW, Simon AH. Cirrhosis of the liver following viral hepatitis, a twenty year mortality follow-up. Am J Epidemiol 1970;92:279-86. �Page 86 Beebe GW, Simon AH, Vivona S. Long-term mortality follow-up of army recruits who received adjuvant influenza virus vaccine in 1951-53. Am J Epidemiol 1972;95:337-46. Berkley MC, Magee KR. Neuropathy following exposure to a dimethylamine salt of 2,4-D. Arch Int Med 1963;111:133-4. Berkman LF, Syme SL. Social networks, host resistance, and mortality: A nine-year follow-up study of Alameda County residents. Am J Epidemiol 1979;109(2):186-204. Binns W, Balls L. Non-teratogenic effects of 2,4,5-trichlorophenoxyacetic acid and 2,4,5-T propylene glycol butyl esters in sheep. Teratology 1971;4(2):245. Bionetics Research Labs of Litton Ind. 1968. Progress report on program of carcinogenesis studies. Vol. 2. Teratogenic study in mice and rats. National Cancer Institute. Bishop YMM, Fienberg SE, Holland PW. Discrete Multivariate Analysis. and Practice. The MIT Press, Cambridge, Mass, 1975. Theory Bjorklund N, Erne K. Toxicological studies of phenoxyacetic herbicides in animals. Acta Vet Scand 1966;7:364-90. Blano EF, Beebe GW. Missiles in the heart. N Engl J Med 1966;274:1039-46. Bleiberg J, Wallen M, Brodkin R, Applebaum IL. Industrially acquired porphyria. Arch Dermatol 1964;89:793-7. Bobowick AR, Kurtzke JF, Brody JA, Hrubec Z, Gilleepie M. Twin study of multiple sclerosis: An epidemiologic inquiry. Neurology 1978;28:978-87. Boeri R, Bordo B, Crenna P, Filippini G, Massetto M, Zecchini A. Preliminary results of a neurological investigation of the population exposed to TCDD in the Seveso region. Riv Pat Nerv Ment 1978;99:111-28. Bogen G. Symptoms in Vietnam veterans exposed to Agent Orange. JAMA 1979; 242:2391 (letter). Boman B. The Vietnam veteran ten years on. Australia & New Zealand J Psychiatry 1982;6(3):107-27. Borus JF. The reentry transition of the Vietnam veteran. Armed Forces & Society 1975;2(1);97-114. Boscarino J. Current excessive drinking among Vietnam veterans: A comparison with other veterans and non-veterans. Internet1 J Psychiatry 1981;27(3):204-12. Boscarino J. Excessive drinking among Vietnam veterans: A possible symptom of post-traumatic stress disorder. Market Opinion Research, Detroit, Michigan. 1980. �Page 87 Boscarino J, Figley CR. Alcohol abuse and post-traumatic stress disorder among Vietnam veterans: A possible link. Market Opinion Research, Detroit, Michigan. 1981. Bower C, Stanley FJ. Herbicides and cleft lip and palate. Lancet 1980;2:1247. Breslow L, Enstrom JE. Persistence of health habits and their relationship to mortality. Preventive Medicine 1980;9:469-83. Breslow NE, Day NE. The analysis of case-control studies. Pub. No. 32, Intl Agency Res on Cancer, 1980, Lyon. 1ARC Scientific Brill NQ, Beebe GW. Follow-up study of psychoneuroses. Am J Psychiatry 1951;108:417-25. Brogan WF, Brogan CE, Dadd JT. Herbicides and cleft lip and palate. 1980;2:597. Lancet Brubaker ML, Binford CH, Trautman JR. Occurrence of leprosy in U.S. veterans after service in endemic areas abroad. Pub Health Rep 1969;84:1051-8. Brunner JA, Brunner GA. Are voluntarily unlisted telephone subscribers really different? J Market Res 1971;8:121-124. Buselmaier MV, Rohrborn G, Propping P. Pesticide mutagenicity investigations by the host mediated assay and the dominant lethal test in mice. Biol Zentralbl 1972;91:311-25. Carter CD, Kimbrough, RD, Liddle JA, Cline RE, Zack MM, Barthe1 WF, Koehler RE, Phillips PE. Tetrachlorodibenzodioxin: An accidental poisoning episode in horse arenas. Science 1975;188(4189):738-40. Case AA, Coffman JR. Waste oil: Toxic for horses. Vet Clin North Am 1973;3(2):273-7. Cederlof R, Friberg L, Hrubec Z. Cardiovascular and respiratory symptoms in relation to tobacco smoking. A study on American twins. Arch Environ Health 1969;18:934-40. Chodoff P. Late effects of concentration camp syndrome. 1963;8:323-33. Arch Gen Psychiatry Christian JC, Feinleib M, Hulley SB, Castell WP, Fabsitz RR, Garrison RJ, Borhani ND, Rosenman RH, Wagner J. Genetics of plasma cholesterol and triglycerides: A study of adult male twins. Acta Genet Med Gemellol 1976;25:145-9. Cleary SF, Pasternak BS, Beebe GW. Cataract incidence in radar workers. Arch Environ Health 1965;11:179-82. Coggon D, Acheson ED. Do phenoxyherbicides cause cancer in man? 1982;!:1057-9. Lancet �Page 88 Cohen BM. Methodology of record follow-up studies on veterans. AJFH 1953;43:1292-8. Collins TFX, Williams CH, Gray GC. Teratogenic studies with 2,4,5-T and 2,4-D in the hamster. Bull Environ Contain Toxicol 1971;6(6):559-67. Cook RR. Dioxin, chloracne and soft tissue sarcoma. Lancet 1981;l:618-9. Cooper SL. Random sampling by telephone: an improved method. J Market Res 1964;1:45-48. Courtney KD, Gaylor DW, Hogan MD, Falk HL, Bates RR, Mitchell I. Teratogenic evaluation of 2,4,5-T. Science 1970;168:864-6. Cox DR. The Analysis of Binary Data, Methuen, 1970, London. Crow KD. Direct testimony before the U.S. Environmental Protection Agency, FIFRA Docket No. 415 et al., Nov. 14, 1980. Cutting RK, Phuoc TH, Ballo JM, Benenson MW, Evans CH. Congenital malformations, hydatidiform moles and stillbirths in the Republic of Vietnam 1960-1969. (Washington, D.C.; Department of Defense, U.S. Government Printing Office No. 903-233). 1970. 29 pp. Dalderup LM. Safety measures for taking down buildings contaminated with toxic material. J Soc Geneesk 1974;52:616-23. Danon JM, Karpati G, Carpenter S. Subacute skeletal myopathy induced by 2,4-dichlorophenoxyacetate in rats and guinea pigs. Muscle and Nerve 1978;(1) Mar/Apr 89-102. Davies IH. Incidence of leukaemia and allied disorders in Western Australian World War II ex-servicemen. Med J Aust 1978;2:321-2. DeBakey ME, Beebe GW. Medical follow-up studies on veterans. JAMA 1962;182:1103-9. DeFazio VJ, Rustin S, Diamond A. Symptom development in Vietnam era veterans. Am J Orthopsychiatry 1975;45(1):158-63. Desi I, Sos J, Olasz J, Sule F, Markus V. Nervous system effects of a chemical herbicide. Arch Environ Health 1962;4:101-8. Donovan JW, Adena MA, Rose G, Battistatta D. Report to the Minister for Veterans' Affairs: Case-Control Study of Congenital Anomalies and Vietnam Service (Birth Defects Study), Australian Government Publishing Service, 1983, Canberra Drill VA, Hiratzska T. Toxicity of 2,4-dichlorophenoxyacetic acid and 2,4,5-trichlorophenoxyacetic acid. Arch Ind Hyg Occ Med 1953;7:61-7. Dugois P, Arablard P, Aimard M, Deshors G. Acne chlorique collective et accidentelle d'un type nouveau. Bull Soc Franc Derm Syph 1968;75:260-1. �Page 89 Dugois P, Marechal J, Colomb L. Acne chlorique au 2,4,5-trichlorophenol. Arch Mai Prof 1958;19:626-7. EPA suspends the major uses of two herbicides. Environ Sci and Tech 1979;13(6):640-1. Eberstein A, Goodgold J. Experimental myotonia induced in denervated muscles by 2,4-D. Muscle and Nerve 1979;2:364-8. Editorial. Phenoxyherbicides, trichlorophenols and soft-tissue sarcomas. Lancet 1981;1:1051-2. Egendorf A, Kadushin C, Laufer R, Rothbart G, Sloan L. Legacies of Vietnam: Comparative adjustment of veterans and their peers. Veterans Administration. 1981. Einsinger FM. Soft tissue tumors. Year Book Medical Publishers 1982. Elsom KA, Beebe GW, Sayen JJ, Scheie HG, Gammon GD, Wood FC. Scrub typhus: follow-up study. Ann Intern Med 1961;55:784-95. Engleman EP, Hollister LE, Kolb FO. 1954; 155 .-1134-40. A Sequelae of rheumatic fever in men. JAMA Eriksson M, Hardell L, Berg N, MoHer T, Avelson 0. Soft-tissue sarcomas and exposure to chemical substances: A case-referent study. Br J Indus Med 1981;38:27-33. Fabsitz RR, Feinleib M, Hrubec Z. Weight changes in adult twins. Med Gemellol 1980;29:273-9. Acta Genet Fabsitz RR, Garrison RJ, Feinleib M, Hjortlano M. A twin analysis of dietary intake. Behav Genet 1978;8:15-25. Faith RE, Luster MI, Moore JA. Chemical separation of helper cell function and delayed hypersensitivity responses. Cell Immune1 1978;40:275-84. Faith RE, Moore JA. Impairment of thymus-dependent immune functions by exposure of the developing immune system to 2,3,7,8-tetrachlorodibenzo-pdioxin. J Toxicol Environ Health 1977;3:451-64. Field B, Kerr C. Herbicide use and incidence of neural-tube defects. Lancet 1979;I(8130):1341-2. Figley CR. The American Legion study of psychological adjustment among Vietnam veterans. Lafayette, Ind. Perdue University. 1977. Figley CR (Ed.). Stress disorders among Vietnam veterans. Brunner/Mazel. 1978. New York. Filippini G, Bordo B, Crenna P, Massetto N, Musicco M, Boeri R. Relationship between clinical and electrophysiological findings and indicators of heavy exposure to 2,3,7,8-tetrachlorodibenzo-dioxin. Scand J Work Environ Health 1981;7:257-62. �Page 90 Fischer V, Boyle JM, Bucuvalas M, Schulraan MA. Myths and realities: A study of attitudes toward Vietnam era veterans. Washington, DC. Louis Harris and Associates, Inc. 1980. Frye JS, Stockton RA. Discriminant analysis of post-traumatic stress disorder among a group of Vietnam veterans. Am J Psychiatry 1982;139(1):52-6. Gaines TB, Holson JF, Nelson CJ, Schumacher HJ. Analysis of strain differences in sensitivity and reproducibility of results in assessing 2,4,5-T teratogenicity in mice. Tox Appl Pharmacol 1974;33:174-5. Giles G, Lickiss JN, Panton J, Baikie MJ, Lowenthal RM. Lymphoma in Tasmania and service in the Armed Forces. (Letter to editor.) Med J Aust 1980;2:339-40. Glasser GJ, Metzger GD. Random digit dialing as a method sampling. J Market Res 1972;9:59-64. of telephone Goldmann PJ. Extremely severe acute chloracne due to trichlorophenol decomposition products. A contribution to the perna problem. Arbeitsmedizin Socialmedizin Arbeitshygiene 1972;7(1):12-8. Goldmann PJ. Severe acute chloracne, a mass intoxication due to 2,3,7,8-tetrachlorodibenzo-dioxin. Der Hausarzt 1973;24(4):149-52. Goldstein JA, Hickman P, Bergman H, Vos JG. Hepatic porphyria induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Res Com Chem Path Pharmacol 1973;6(3):919-28. Goldstein NP, Jones PH, Brown JR. Peripheral neuropathy after exposure to an ester of dichlorophenoxyacetic acid. JAMA 1959;171(10):1306-9. Cover KR, McEaddy BJ. Job situation of Vietnam-era veterans. Review 1974;96(8):17-26. Monthly Labor Greene MH. Non-Hodgkins lymphomas. In: Schottenfeld D, Fraumeni JF, eds. Cancer epidemiology and prevention. Philadelphia: WB Saunders, 1982:754-78. Greig JB. The toxicology of 2,3,7,8-tetrachlorodibenzo-p-dioxin and its structural analogues. Ann Occup Hyg 1979;22:411-20. Grufferman S. Hodgkin's disease. In: Schottenfeld D, Fraumeni JF, eds. Cancer epidemiology and prevention. Philadelphia: WB Saunders, 1982:739-53. Hall SM. Effects on pregnant rats and their progeny of adequate low protein diets containing 2,4,5-trichlorophenoxyacetic acid (2,4,5-T or p,p'-DDT). Fed Proc, Fed Am Soc Exp Biol 1972;31:726. Hammond R. National Survey of Veterans. Government Printing Office. 1980. Veterans Administration. U.S. Hanify JA, Metcalf P, Nobbs CL, Worsley KJ. Aerial spraying of 2,4,5-T and human birth malformations: An epidemiological investigation. Science 1981;212:349-51. �Page 91 Hardell L, Eriksson M, Leaner P, Lungren E. Malignant lymphoma and exposure to chemicals, especially organic solvents, chlorophenols and phenoxy acids: a case-control study. Br J Cancer 1981;43:169-76. Hardell L, Johansson B, Axelson 0. Epidemiological study of nasal and nasopharyngeal cancer and their relation to phenoxy acid or chlorophenol exposure. Am J Ind Med 1982;3:247-57. Hardell L, Sandstrom A. Case-control study: Soft-tissue sarcomas and exposure to phenoxyacetic acids or chlorophenols. Br J Cancer 1979;39:7H-7. Harris and Associates. A study of the problems facing Vietnam era veterans: Their adjustment to civilian life. October 1971. Harris MW, Moore JA, Vos JG, Gupta BN. General biological effects of TCDD in laboratory animals. Environ Health Perspect 1973;5:101-9. Hauck M, Cox M. Locating a sample by random digit dialing. Pub Opin Quart 1976;38:253-260. Havlik RJ, Garrison RJ, Fabsitz R, Feinleib M. Variability of heart rate in twins. J Electrocardiol 1980;13:45-8. Hawryzluk 0. Chronic disease patterns in United States Army officers. Military Medicine 1975;140:89-93. Hay A. Toxic cloud over Seveso. Nature 1976;262:636-8. Helzer JE, Robins LN, Wish E, Hesselbrock M. Depression in Vietnam veterans and civilian controls. Am J Psychiatry 1979;136(4B):526-9. Herbicide Assessment Commission. The American Association for the Advancement of Science. 1970. Summary of presentations. Chicago, Illinois. Hill EV, Carlisle H. Toxicity of 2,4-dichlorophenoxyacetic acid for experimental animals. J Indust Hyg and Toxicol 1947;29(2):85-95. Hoffer A, Pollin W. Schizophrenia in the NAS-NRC panel of 15,909 veteran twin pairs. Arch Gen Psychiatry 1970;23:469-77. Holden C. Agent Orange furor continues to build. Science 1979; 205:770-772* Honchar P. Health hazard evaluation determination. NIOSH. Honchar PA, Halperin WE. Lancet 1981;1:268-9. Report 80-039. 1982. 2,4,5-T, trichlorophenol, and soft tissue sarcoma. Hook EB, Regal RR. Validity of Bunoulki census, log-linear, and truncated binomial models for correcting underestimates in prevalence studies. Am J Epid 1982;116:168-76 Hoover RN, Strasser PH. Artificial sweeteners and human bladder cancer preliminary results. Lancet 1981;1:837-40. �Page 92 Horn JM, Plomin R, Rosenman R. Heritability of personality traits in adult male twins. Behav Genet 1976;6:17-30. Horowitz MJ, Solomon GF. A prediction of delayed stress response syndromes in Vietnam veterans. J Social Issues 1975;31(4)67-80. Hrubec Z, Cederlof R, Friberg L. Respiratory symptoms in twins: Effects of residence-associated air pollution, tobacco and alcohol use, and other factors. Arch Environ Health 1973;27:189-95. Hrubec Z, Nashold BS. Epidemiology of lumbar disc lesions in the military in World War II. Am J Epidemiol 1975; 102:366-76. Hrubec Z, Neel JV. Familial factors in early deaths: Twins followed 30 years to ages 51-61 in 1978. Hum Genet 1981;59:39-46. Hrubec Z, Omen GS. Evidence of genetic predisposition to alcoholic cirrhosis and psychosis. Am J Hum Genet 1980;32:112A (Abstract). Hrubec Z, Ryder RA. Traumatic limb amputations and subsequent mortality from cardiovascular disease and other causes. J Chron Dis 1980;33:239-50. Hrubec Z, Zukel WJ. Epidemiology of coronary heart disease among young army males of World War II. Am Heart J 1974;87:722-30. Huff JE, Moore JA, Saracci R, Toraatis L. Long-term hazards of polychlorinated dibenzodioxins and polychlorinated dibenzofurans. Environ Health Perspect 1980;36:221-40. Ideo G, Bellati G, Bellobuono A, Mocarelli P, Marocchi A, Brambilla P. Increased urinary d-glucaric acid excretion by children living in an area polluted with tetrachlorodibenzoparadioxin (TCDD). Clinica Chimica Acta 1982;120:273-83. International Agency for Research on Cancer, IARC. Monographs on the evaluation of the carcinogenic risk of chemicals to man: Some fumigants, the herbicides 2,4-D and 2,4,5-T, chlorinated dibenzodioxins and miscellaneous industrial chemicals. Vol 15 pp 41-103, 1977. JabIon S, Miller RW. Army technologists: 29-year followup for cause of death. Radiology 1978;126:677-9. Jensen NE. Chloracne: 3 cases. Proc Roy Soc Med 1972;65:21-2. Jirasek L, Kalensky J, Kubec K. Acne chlorina and porphyria cutanea tarda during the manufacture of herbicides. Part I. Cesk Dermatol 1973;48(5):306-15. Jirasek L, Kalensky J, Kubec K, Pazderova J, and Lukas E. Acne chlorina, porphyria cutanea tarda and other manifestations of general intoxication during the manufacture of herbicides. Part II. Cesk Dermatol 1974;49(3):145-57. Johnson FE, Kugler MA, Brown SM. Soft-tissue sarcomas and chlorinated phenols. Lancet 1981;2:40. �Page 93 Keehn RJ. Follow-up studies of World War II and Korean conflict prisoners. Am J Epidemiol 1980;111:194-211. Keehn RJ. Military rank at separation and mortality. Armed For Soc 1978;4:283-92. Keehn RJ, Goldberg ID, Beebe GW. Twenty-four years mortality followup of army veterans with disabling separations for psychoneurosis in 1944. Psychosom Med 1974;101:27-46. Kimbrough RD. 2,3,7,8-tetrachlorodibenzodioxin (TCDD) - toxicity in animals relevance to human health, with notes on 2,4,5-T picloram, cacodylic acid, and 2,4-D. In Proceedings from the 2nd Continuing Education Conference on Herbicide Orange, Washington, D.C. May 28-30, 1980. Kimbrough RD, Carter CD, Liddle JA, Cline RE, Phillips PE. Epidemiology and pathology of a tetrachlorodibenzodioxin poisoning episode. Arch Environ Health 1977;32(2):77-86. Kimmig J, Schulz KH. Berufliche akne durch chlorierte aromatische zyklische ather. Dermatologica 1957;115:540-6. Klecka W, Tuchfarber AJ Jr. Method effects on behavioral and attitudinal measures: Random digit dialing compared to personal surveys. Univ Cincinnati, Behavioral Sciences Lab, 1976, 9p. Presented at the annual meeting of the Midwest Association for Public Opinion Research, Chicago, Illinois, November 19-20, 1976. Klecka W, Tuchfarber AJ Jr. Random digit dialing as an efficient method for political polling. Georgia Polit Assoc J 1974a;2:133-151. Klecka W, Tuchfarber AJ Jr. The efficiency, biases, and problems of random digit dialing. Univ Cincinnati, Behavioral Sciences Lab., 1974b, 30p. Presented at the Directors Conference, Temple University, June 1974 and the annual meeting of the American Association for Public Opinion Research, Bolton Landing, New York, May 31-June 2, 1974. Klecka W, Tuchfarber AJ Jr. The efficacy of random digit dialing, Survey Research Newsletter, University of Illinois, Survey Research Lab 1973;5:14-15. Klonoff H, McDougall G, Clark C, Kramer P, Horgan J. The neurological psychiatric, and physical effects of prolonged and severe stress: 30 years later. J Nerv Ment Dis 1976;163:246-7. Kociba RJ, Keeler PA, Park CN, Gehring PJ. 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD): Results of a 13-week oral toxicity study in rats. Toxicol Appl Pharmacol 1976;35:553-74. Kociba RJ, Keyes DG, Beyer JE, et al. Results of a two year chronic toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-dioxin in rats. Toxicol Appl Pharmacol 1978;46:279-303. �Page 94 Kociba RJ, Schwetz BA. A review of the toxicity of 2,3,7,8-tetrachlorodibenzop-dioxin (TCDD) with a comparison to the toxicity of other chlorinated dioxin isomers. Quarterly Bull Assn Food & Drug Officials 1982;46:168-88. Kontek M, Jasinski K, Marcinkowska B, Tokarz F, Pietraszek Z, Handschuh R. Electroencephalographic study of farm workers exposed to derivatives of arylalcanocarboxylic acids. Polski Tygodnik Lekarski 1973;28(25):937-9. Kurtzke JF, Beebe GW. Epidemiology of amyotrophic lateral sclerosis. Neurology 1980;30:453-62. Lamb JC, Moore JA, Marks TA. Evaluation of 2,4-dichlorophenoxyacetic acid (2,4-D), 2,4,5-trichlorophenoxyacetic acid (2,4,5-T), and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) toxicity in C57BL/6 mice: Reproduction and fertility in treated male mice and evaluation of congenital malformations in their offspring. National Toxicology Program, Research Triangle Institute, Research Triangle Park, NC. Report No. NTP-80-44. 57 pp. 1980. Layde PM, Webster LA, Wingo PA, Schlesselman JJ, Dry HW, and the Cancer and Steroid Hormone Study Group. Long-term oral contraceptive use and the risk of breast cancer. JAMA 1983;249:1591-5. Lipkin JO, Blank AS, Parson ER, Smith J. Vietnam veterans and post-traumatic stress disorder. Hospital and Community Psychiatry 1982;33(11):908-12. Luster MI, Faith RE, Clark G. Laboratory studies on the immune effects of halogenated aromatics. Ann NY Acad Sci 1979;320:473-86. Lynfield YL. Skin diseases in twins. Arch Dermatol 1974;110:722-4. Malizia E, Andreucci G, Chiaverelli M, Amato A, Gagliardi L. A follow-up of 20 months of Seveso, an environmental calamity. Vet Hum Toxicol 1979;21(Suppl):136-40. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Can Inst 1959;22:719-48. Mantel N. An evaluation of the statistical methods used in EPA's "Report of assessment of a field investigation of six-year spontaneous abortion rates in three Oregon practices" (the ALSEA II Report). Unpublished. Bethesda, Maryland: Biostatistics Center, George Washington University. 1979. Mantovani A, Vecchi A, Luini W, Sironi M, Candioni GP, Spreafico, and Garratini S. Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on macrophage and natural killer cell mediated cytotoxicity in mice. Institute di Ricerche Farmacologiche "mario negri" via Enitrea 62-2, Milano, Italy. 1979. Martindale M, Poston DL. Variations in veteran and nonveteran earnings patterns among World War II, Korea and Vietnam war cohorts. Armed Forces in Society 1979;5(2):219-43. Maskin M. Psychological stress and the coping process. York, 1966. McGraw-Hill, New �Page 95 May G. Chloracne from the accidental production of tetrachlorodibenzodioxin. Brit J Industr Med 1973;30:276-83. May G. Tetrachlorodibenzodioxin: A survey of subjects ten years after exposure. Brit J Industr Med 1982;39:128-35. McConnell EE, Moore JA, Dalgard DW. Toxicity of 2,3,7,8-tetrachlorodibenzo-pdioxin in Rhesus monkeys (Macaca mulatta) following a single oral dose. Toxicol Appl Pharmacol 1978(a);43(1):175-87. McConnell EE, Moore JA, Haseman JK, Harris MW. The comparative toxicity of chlorinated dibenzo-p-dioxins in mice and guinea pigs. Toxicol Appl Pharmacol 1978(b);44(2):335-56. McQueen EG, Veale AMO, Alexander WS, Bates MM. 2,4,5-T and human birth defects. Report prepared by Dept of Health, New Zealand. 1977. 41 pp. Meselson MS, Westing AH, Constable JD. Background meterial relevant to presentations at the 1970 annual meeting of the AAAS concerning the Herbicide Assessment Commission for the American Association for the Advancement of Science. Washington, D.C. Min., 1971, 47 p. Miller RW, Beebe GW. Infectious mononucleosis and the empirical risk of cancer. JNCI 1973;50:315-21. Miller RW, JabIon S. A search for late radiation effects among men who served as x-ray technologists in the U.S. Army during World War II. Radiology 1970;96:269-74. Milligan WL, Powell DA. Learning and reaction time performance in older veterans: Relationship to attitudes and life satisfaction. Int J Aging Human Dev 1981;13:151-68. Mintz J, O'Brien CP, Pomerantz B. The impact of Vietnam service on heroin-adjusted veterans. Am J Drug & Alcohol Abuse 1979;6(l):39-52. Moses M, Selikoff IJ. Soft tissue sarcomas, phenoxyherbicides and chlorinated phenols. Lancet 1981;1:1370. Nace EP, O'Brien CP, Mintz J, Meyers AL, Ream N. Follow-up of Vietnam veterans. II. Social adjustment. Drug and Alcohol Depending 1980;6(4):209-14. National Academy of Science. Committee on the effects of herbicides in South Vietnam. Part A. Summary & conclusions. Washington, D.C., 1974, 398 pp. National Cancer Institute DHHS Publication No. NIH 80-1765. National Cancer Institute. Surveillance epidemiology end results, Incidence and mortality data: 1973-77. National Cancer Institute. Monograph 57, June 1981. National Toxicology Program. Carcinogenesis bioassay of 2,3,7,8 tetrachlorodibenzo-p-dioxin (CAS Nol 1746-01-6) in Swiss-Webster mice (dermal study). Natl Toxicol Program Tech Rep Ser 1982a;Issue 201:113pp. �Page 96 National Toxicology Program. Carcinogenesis bioassay of 2,3,7,8 tetrachlorodibenzo-p-dioxin (CAS Nol 1746-01-6) in Osborne-Mendel rats and B6C3F1 mice (gavage study). Natl Toxicol Program Tech Rep Ser 1982b;Issue 209:195pp. Neel S. Medical support of the U.S. Army in Vietnam, 1965-1970. United States Army. Washington, D.C., 1973. Nefzger MD. Follow-up studies of World War II and Korean War prisoners. Am J Epidemiol 1970;91:123-38. Nelson CJ, Holson JF, Green HG, Gaylor DW. Retrospective study of the relationship between agricultural use of 2,4,5-T and cleft palate occurrence in Arkansas. Teratology 1979;19:377-384. Norman JE. Lung cancer in World War I veterans with mustard-gas injury: 1919-1965. JNCI 1975;54:311-7. Norman JE, Robinette CD, Fraumeni JF. The mortality experience of Army World War II chemical processing companies. J Occup Med 1981;23:818-22« Nostrom A, Rappe C, Lindahl R, Buser HR. Analysis of some older Scandinavian formulations of 2,4-dichlorophenoxy acetic acid and 2,4,5-trichlorophenoxy acetic acid for contents of chlorinated dibenzo-p dioxins and dibenzofurans. Scand J Work Environ and Health 1979;5:375-378. O'Brien CP, Nace EP, Mintz J, Meyers AL, Ream N. Follow-up of Vietnam veterans. I. Relapse to drug use after Vietnam service. Drug and Alcohol Depending 1980;5(5):333-40. Oglesby P (edit). Studies of hypertension in twins. Second International Symposium on the Epidemiology of Hypertension. Miami, 1975. Ognibene AJ, Barrett 0 (edit). Internal Medicine in Vietnam, Vol. II. General medicine and infectious diseases. United States Army. Washington, D.C., 1982. Oliver RM. Toxic effects of 2,3,7,8 tetrachlorodibenzo 1,4 dioxin in laboratory workers. Br J Ind Med 1975;32:49-53. Olson JR, Holscher MA, Neal RA. Toxicity of 2,3,7,8-tetrachlorodibenzo-pdioxin in the golden Syrian hamster. Toxicol Appl Pharmacol 1980;55:67-78. Pazderova-Vejlupkova J, Lukas E, Nemcova M, Pickova J, Jirasek L. Chronic poisoning by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Pracov Lek 1980;32:204-209. Pazderova-Vejlupkova J, Nemcova M, Pickova J, Jirasek L, Lukas E. The development and prognosis of chronic intoxication by tetrachlorodibenzo-pdioxin in men. Arch Environ Health 1981;36:5-11. Penk WE, Robinowitz R, Roberts WR, Patterson ET, Dolon MP, Atkins HG. Adjustment differences among male substance abusers varying in degree of combat experience in Vietnam. J Consulting & Clinical Psychology 1981;49(3):426-37. �Page 97 Pitot HC, Goldsworthy T, Campbell HA, Poland A. Quantitative evaluation of the promotion by 2,3,7,8-tetrachlorodibenzo-p-dioxin of hepatocarcinogenesis from diethylnitrosamine. Cancer Res. 1980;40:3616-20. Pocchiari F, Silano V, Zampieri A. Human health effects from accidental release of tetrachlorodibenzo-p-dioxin (TCDD) at Seveso, Italy. Ann NY Acad Sci 1979;320:311-320. Poland AP, Smith D, Metter G, Fossick P. A health survey of workers in a 2,4-D and 2,4,5-T plant. Arch Environ Health 1971;22:316-327. Pollin W, Allen MG, Hoffer A, Stabenau JR, Hrubec Z. Psychopathology in 15,909 pairs of veteran twins. Am J Psychiatry 1969;126:597-610. Puri ML, Sen PK. Nonparametric Methods in Multivariate Analysis. & Sons (Eds.), 1971, New York. John Wiley Reggiani G. Acute human exposure to TCDD in Seveso, Italy. J Toxicol Environ Health 1980;6:27-43. Reggiani G. Estimation of the TCDD toxic potential in the light of the Seveso accident. Arch Toxicol 1979;2:291-302. Roberts WR, Penk WE, Gearing ML, Robinowitz R, Dolan MP, Patterson ET. Interpersonal problems of Vietnam combat veterans with symptoms of post traumatic stress disorder. J Abnormal Psychology 1982;91(6):444-50. Robinette CD. Splenectomy and subsequent mortality in veterans of the 1939-45 war. Lancet 1977;2:127-8. Robinette CD, Fraumeni JF. Asthma and subsequent mortality in World War II veterans. J Chron Dis 1978;31:619-24. Rogers EL, Goldkind L, Goldkind SF. Increasing frequency of esophageal cancer among black male veterans. Cancer 1982;49:610-7. Rogot E, Murray JL. Smoking and causes of death among U.S. veterans: years of observation. Pub Health Rep 1980;95:213-22. 16 Rose HA, Rose SP. Chemical spraying as reported by refugees from South Vietnam. Science 1972;177(4050):710-2. SRI International. A case-control study of the relationship between exposure to 2,4-D and spontaneous abortion in humans. National Forest Products Association, Washington, D.C. 1981. 116 p. Sarma PR, Jacobs J. Thoracic soft-tissue sarcoma in Vietnam veterans exposed to Agent Orange. NEJM 1982;306:1109. Scheffe H. York. The Analysis of Variance. John Wiley & Sons (Eds.), 1959, New Schulz KH. Klinische und experimentelle untersuchungen zur atiologie der chloracne. Arch Klin Exp Dermatol 1957;206:589-96. �Page 98 Schwartz JT, Reviling FH, Feinleib M, Garrison RJ, Collie DJ. Twin heritability study of the effect of corticosteroids on intraocular pressure. J Med Genet 1972;9:137-43. Schwartz JT, Reuling FH, Feinleib M, Garrison RJ, Collie DJ. Twin heritability study of corticosteroid response. Trans Am Acad Ophthalmol Otolaryngol 1973:77:126-36. Schwartz JT, Reuling FH, Feinleib M, Garrison RJ, Collie DJ. Twin study on ocular pressure after topical dexaraethasone. Am J Ophthalmol 1973;76:126-36. Schwartz JT, Reuling FH, Feinleib M, Garrison RJ, Collie DJ. Twin study on ocular pressure following topically applied dexamethasone. Arch Ophthalmol 1973;90:281-86. Segal DR. Illicit drug use in the U.S. Army. 1977;18:66-83. Sociological Symposium Seltzer CC, JabIon S. Army rank and subsequent mortality by cause: follow-up. Am J Epidemiol 1977;105:559-66. 23 year Seltzer CC, JabIon S. Effects of selection on mortality. Am J Epidemiol 1974;100:367-71. Sinclair PR, Granick S. Uroporphyrin formation induced by chlorinated hydrocarbons (lindane, polychlorinated biphenyls, tetrachlorodiobenzo-pdioxin). Requirements for endogenous iron, protein synthesis and drug-metabolizing activity. Biochem Biophys Res Coraraun 1974;61:124-33. Singer R, Moses M, Valciukas J, Lilis R, Selikoff IJ. Nerve conduction velocity studies of workers employed in the manufacture of phenoxy herbicides. Environ Res 1982;29:297-311. Smith AH, et al. Preliminary report of reproductive outcomes among pesticide applicators using 2,4,5-T. New Zealand Med J 1981;93:177-9. Sparschu GL, Dunn FL, Rowe VK. Study of the teratdgenicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the rat. Fd Cosmet Toxicol 1971;9:405-12. Stabenau JR, Pollin W, Allen MG. Twin studies in schizophrenia and neurosis* Semin Psychiatry 1970;2:65-74. Starr P, Henry J, Bonner R. The discarded army: Veterans after Vietnam. York. Charterhouse, 1973. New Stellman S, Stellman J. Health problems among 535 Vietnam veterans potentially exposed to toxic herbicides. Abstract. Am J Epidemiol 1980;112(3):444. Stinson JF. Vietnam veterans in the labor market of the 1970's. Monthly Labor Review 1979;48(4):3-ll. Strik JJTWA. Porphyrins in urine as an indication of exposure to chlorinated hydrocarbons. Ann NY Acad Sci 1979;320:308-10. �Page 99 Suskind RR. Chloracne and associated health problems in the manufacture of 2,4,5-T. Report to the Joint Conference, National Institute of Environmental Health Sciences and International Agency for Research on Cancer, WHO, Lyon, France, January 11. 1978. 7 pp. Schwetz BA, Norris JM, Sparschu GL, Rowe VK, Gehring PJ, Emerson JL, Gerbig CG. Toxicity of chlorinated dibenzo-p-dioxins. Environ Health Perspect 1973;5:87-99. Taubman P. The determinants of earnings: Economic Rev 1976;66:858-70. Taylor JS. Environmental chloracne: 1979;320:295-307. A study of white male twins. Am Update and overview. Ann NY Acad Sci Telegina KA, Bikbulatova LI. Affection of the follicular apparatus of the skin in workers occupied in production of butyl ether of 2,4,5-trichlorophenoxyacetic acid. Vestn Dermatol Venerol 1970;44:35-9. Texas Department of Health. Texas Veterans Agent Orange Assistance Program. Annual Report. March 31, 1983. Austin, Texas. Thiess AM, Frentzel-Beyme R, Link R. Mortality study of persons exposed to dioxin in a trichlorophenol-process accident that occurred in the BASFAG on November 17, 1953. Am J Indus Med 1982;3:179-189. Thigpen JE, Faith RE, McConnell EE, Moore JA. Increased susceptibility to bacterial infection as a sequela of exposure to 2,3,7,8-tetrachlorodibenzo-pdioxin. Infection and Immunity 1975;12(6):1319-24. Toth K, Samfaie-Relle S, Sugar J, Bence J. Carcinogenicity testing of the herbicide 2,4,5-trichlorophenoxy ethanol containing dioxin and pure dioxin in Swiss mice. Nature 1979;278:548-9. Townsend JC, Bodner KM, Van Peenen PFD, Olson RD, Cook RR. Survey of reproductive events of wives of employees exposed to chlorinated dioxins. J Epidemiol 1982;115:695-713. Am Troldahl VC, Carter RE. Random selection of respondents within households in phone surveys. J Market Res 1964;1:71-76. Tuchfarber AJ, Klecka W. Demographic similarities between samples collected by random digit dialing versus complex sampling designs. Univ Cincinnati, Behavioral Sciences Lab. Presented at the annual meeting of the American Association for Public Opinion Research, Itasca, Illinois, May 29-June 1, 1975. Tucker MA, Fraumeni JF. Soft tissue. In: Cancer epidemiology and prevention. Schottenfeld D, Fraumeni JF, eds. Philadelphia: WB Saunders, 1982;827-836. Tung TT, Anh TK, Tuyen BQ, Tra DX, Huyen NX. Clinical effects of massive and continuous utilization of defoliants on civilians. Vietnamese Studies 1971;29:53-81. �Page 100 UCLA Study Protocol Prepared for the Veterans Administration (Spivey GH, Detels R), Epidemiologic Studies of Agent Orange. January 22, 1982. U.S. Army. Recent advances in medicine and surgery: Japan and Korea (presentations). Vol. II. Army Medical Service Graduate School. Washington, D.C, 1954. Veterans Administration. Review of literature on herbicides, including phenoxy herbicides and associated dioxins. Volume 1. Analysis of literature. Washington, D.C.: U.S. Government Printing Office, October 1981-0-522-609/21. Veterans Administration. Review of literature on herbicides, including phenoxy herbicides and associated dioxins. Volume II. Annotated bibliography. Washington, D.C.: U.S. Government Printing Office, October 1981-0-522-610/22. Vos JG, Kreeftenberg JG, Engel HWB, Minderhoud A, Van Noorle Jansen LM. Studies on 2,3,7,8-tetrachlorodibenzo-p-dioxin induced immune suppression and decreased resistance to infection: Endotoxin hypersensitivity, serum zinc concentrations and effect of thyraosin treatment. Toxicology 1978;9:75-86. Vos JG, Moore JA. Suppression of cellular immunity in rats and mice by maternal treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Int Arch Allerg Appl Immunol 1974;47:777-94. Vos JG, Moore JA, Zinkl JG. Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on the immune system of laboratory animals. Environ Health Perspect 1973;5:149-62. Wade N. Viets and vets fear herbicide health effects. Science 1979; 204:817. V Wagner SL, Witt JM, Norris LA, Higgins JE, Agresti A, Ortiz M. A scientific critique of the EPA Alsea II study and report. Environmental Health Sciences Center, Oregon State University, Corvallis, Oregon. 1979. Waksberg J. Sampling methods for random digit dialing. 1978;73:40-46. J Amer Stat Assoc Walker AE, Martin JV. Lipid profiles in dioxin-exposed workers. Lancet 1979;l:446-7. Wallis GG. Stress in service families. Proc Royal Soc Med 1968;61:976-8, Wallis WE, Van Poznak A, Plum F. Generalized muscular stiffness, fasciculations, and myokymia of peripheral nerve origin. Arch Neurol 1970;22:430-9. Weller T. Telephone interviewing procedures. Survey Research Newsletter, University of Illinois, Survey Research Lab 1973;5:13-14. Wilson JG. Abnormalities of intrauterine development in non-human primates. In Diczfalusy E, and Standley CC, eds. The use of non-human primates in research on human reproduction. Acta Endocrin 1971;166(Suppl):261-92. �Page 101 Wilson JP, Krauss GE. Predicting post-traumatic stress syndromes among Vietnam veterans. Paper presented at the 25th Neuropsychiatric Institute, VA Medical Center, Coatsville, PA. October 21, 1982. Wilson JP. Towards an understanding of post-traumatic stress disorder among Vietnam veterans. Testimony before U.S. Senate Subcommittee on Veterans Affairs. May 21, 1980. Wolfe WH. Human health effects following exposure to the phenoxy herbicides and TCDD. In Veterans Administration - Proceedings from the 2nd Continuing Education Conference on Herbicide Orange. Washington, D.C. May 28-30, 1980. Young AL, Cacagni JA, Thalken CE, Tremblay JW. The Toxicology, Environmental Fate, and Human Risk of Herbicide Orange and Its Associated Dioxin. USAF OEHL TR-78-92. USAF Occupational and Environmental Health Laboratory, Aerospace Medical Division, Brooks Air Force Base, Texas, 1978. Zack JA, Suskind RR. The mortality experience of workers exposed to tetrachlorodibenzodioxin in a trichlorophenol process accident. J Occup Med 1980;22:ll-4. Ziegler OK, Hassanein RS, Harris 0, Stewart R. Headache in a non-clinic twin population. Headache 1975;14:213-8. � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 064 Folder The folder containing the original item. 1728 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Description An account of the resource <strong>Corporate Author: </strong>Centers for Disease Control, Public Health Service, U.S. Department of Health and Human Services, Atlanta, Georgia Date A point or period of time associated with an event in the lifecycle of the resource 1983-11-01 Title A name given to the resource Protocol for Epidemiologic Studies of the Health of Vietnam Veterans Subject The topic of the resource Agent Orange Exposure Study Vietnam Experience Study Selected Cancers Study study protocol ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/f87437ff884ffa0584af30af6ac22c94.pdf a1518b40e7e8cb96f6a95ee2fa0e411c PDF Text Text Item ID Number 01729 Author Corporate Author Centers for Disease Control Roport/Artido TitiB Supporting Statement: Epldemlologlc Study of the Health of Vietnam Veterans Journal/Book Title Year 1983 Month/Day November Color Number of Imaoas n 31 Dascplpton Notes Monday, June 11, 2001 Page 1730 of 1793 �Supporting Statement Epidemiclogic Study of the Haalth of Vietnam Veterans Centers for Disease Control November 1983 �t *>' , Contents I. Supporting Statement A. Justification 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. B. Background Purpose Information Technology Identification of Duplication Use of Existing Data Small Business Consequences of Less Frequent Collection 5 CFR 1320.6. Consultation Confidentiality Assurances Sensitive Data Cost Respondent Burden Changes in Burden Project Schedule Collection of Information Employing Statistical Methods 1. 2. 3. 4. 5. Respondent Universe Data Collection Procedures Response Rates and Power Pretests and Pilot Studies Statistical Designers II. Attachments 1. 2. 3. 4. 5. 6. 7. 8. Public Law 96-151 Puhlic Law 97-72 Study Protocol (Draft) Confidentiality Assurances Statement A.O./V.E. Questionnaire (Testable Draft) Initial Contact Letter (A.O./V.E.) Initial Telephone Contact (Draft text) Special Cancer Study Questionnaire �Supporting Statement A. Justification 1. Background During the past several years, a large number of Vietnam veterans have coire to believe that they have an unusually high frequency of certain illnesses. Much of their concern steins from presumed exposure to Agent Orange, and to dioxin, a contaminant present in Agent Orange. (Dioxin has been demonstrated to be carcinogenic and teratogenic in laboratory animals). In addition to cancer, veterans have complained of other adverse health effects including neurologic disorders, reproductive problems, and infections. Unfortunately, there is little objective evidence regarding the health of Vietnam veterans relative to other men of similar age. In recognition of this lack of information, Public Law 96-151 (Attachment 1) required that the Veterans Administration (VA) conduct an "epidemiologic" study of U.S. veterans to assess the possible health effects of exposure to herbicides and dioxin during the Vietnam War. Public Law 97-72 (Attachment 2) expanded this mandate to include the study of other environmental exposures which may have occurred in Vietnam. In January, 1S83, the responsibility for design, conduct, and analysis of studies responsive to these laws WMS transferred from VA to CDC by an Interagency Agreement. 2. Purpose A rcnjor concern of Vietnam veterans is that they are at high risk for a variety of diseases. The cause of this putative high risk is generally suspected to he exposure to Agent Orange and other herbicides, hut there js aJso concern that there may have been other factors incidental to Vietnam service which conferred an increased risk. Collection of necessary information <-md performance of CDC's proposed studies should permit an assessment of the validity of both the general and some of the specific concerns. Without these studies the veterans' concerns cannot be addressed and the Congressional mandate cannot be fulfilled. This submission is for an "Agent Orange" (A.O.) Study, a "Vietnam Experience" (V.E.) Study, and a "Selected Cancers" Study (Attachment 3, Protocols). The Agent Orange study is a cohort study designed to try to determine whether or not the. health experience of Vietnam veterans exposed to Agent Orange differs significantly from that of Vietnam veterans not so exposed. Ihis study will involve three cohorts of some 6000 men each. Two of the cohorts wi-11 be drawn from a random sample of combat battalions which served in the III Corps tactical area of Vietnam in 1967-68. This location and time represent an area and period of heavy Agent Orange use. A third cohort will be selected by a different method. Areas for which there is no evidence of herbicide use prior to 1969 will be identified and a list of units which served only in those areas during 1967-68 will be compiled. From this list a sample of units will be drawn and subjects for the third cohort will be selected from that sample. The Vietnam Experience Study will involve two cohorts of 6000 subjects each. It is designed to evaluate whether veterans who served in Vietnam are at greater risk for certain adverse health outcomes than are their counterparts who served elsewhere. Selection of subjects will be based on review of systematically chosen personnel records located at the St. louis records center. Both cohorts will consist of Army first term enlistees or �draftees who served in the non-officer ranks between 1966 and 1971. The first cohort will be those selected individuals who served only in the U.S. end Vietnam; the second cohort will be comprised of three groups: 1) service in U.S. only; 2) service in U.S. and Europe; 3) service in U.S. and Korea. A random subset from each of the five study cohorts will be selected to participate in medical, psychological and laboratory evaluation. The goal will be to complete examinations on 2000 men per cohort. The Selected Cancers Study (SCS) is a population based case control study designed to determine whether men who served in Vietnam are at increased risk of developing soft tissue sarcoma, lymphoma, liver, and nasopharyngeal cancer. Cases will be men with birthdates 1933-53 and identified with one of the selected cancers from July ], 1984 to June 30, J988. Cases will be identified and interviewed by cooperating Surveillance, Epidemiology, and End Result (SEER) Centers. These centers are population-based cancer registries sponsored by the National Cancer Institute (NCI). Controls will be selected by the random digit dialing method and matched to cases by age, sex, and race. Interview of controls will also be performed by the SEER Centers. Subjects in all three studies will be interviewed to collect pertinent information. For the Agent Orange and Vietnam experience studies, this will include: sociodemographic data, medical history, environmental and occupational exposure information, and military history. In addition to these elements, participants in the Selected Cancers Study will be questioned regarding family history of cancer. A.O. and V.E. study interviews will be conducted by telephone, and will be supplemented by in-person contact should pilot testing indicate that participation is suffering because too few study subjects can be reached by telephone. All SCS interviews will be conducted face to face. Estimated response time for each interview is one hour per subject. 3. Information Technology CDC proposes that a "computer assisted telephone interviewing" (CATI) system be employed. The quality control advantages provided by such a system serve to reduce respondent burden by speeding administration of the questionnaire and eliminating call-backs because of interviewer failure. Furthermore, by its nature telephone interviewing is less intrusive than is an in-person technique. The in-person technique was chosen for the Selected Cancers Study because the SEER Centers are familiar with that approach and the logistics of establishing a CATI system in multiple sites are unmanageable. Such an approach would also be extremely expensive. ft. Identification of Duplication The United States Air Force has recently completed a cohort study of the air crews and support personnel involved in aerial spraying of Agent Orange in Vietnam ("Operation Ranch Hand"). This study will provide extensive data regarding health effects resulting from exposure to Agent Orange. However, neither the exposures nor the personnel involved in the Air Force study are representative of the ground forces which are the focus of CDC's proposed studies. Furthermore the Air Force study made no attempt to investigate the health effects of the general "Vietnam Experience." In December, 1980, the Veterans Administration (VA) contracted for a "Review of Literature on Herbicides, Including Phenoxy Herbicides and Associated Dioxins." The report on that review was delivered to VA in September, 1981. Volume I of that report ("Analysis of Literature") stated that a gap in existing information existed in that: "Human health effects from use of defoliants in Vietnam have not been systematically documented." The Agent Orange Study proposed by CDC is designed to help close this gap. �The questions of adverse health effects stemming from the general experience of service in Vietnam as well as possible excesses of soft tissue sarcoma and lymphoroa among Vietnam veterans have not been previously addressed in a scientifically rigorous fashion. The "Selected Cancers" and "Vietnam Experience" studies proposed by CDC represent the first efforts to answer these questions and fulfill the Congressional mandate to conduct epidendologic studies of environmental exposures which may have ocurred in Vietnam. 5. Use of Existing Data Existing data cannot be modified to completely satisfy the requirements of the Agent Orange, Vietnam Experience, and Selected Cancers studies. In the case of the Soft Tissue Sarcoma and Vietnam Experience studies, as previously stated, the data simply do not exist. Regarding the Agent Orange Study, most of the data bearing on this question are drawn from occupational settings, and most reports and studies of workers exposed to TCDD (dioxin) are descriptive. Additionally the age/race/ethnic composition of the groups of factory workers exposed is not comparable to that of American ground forces in Vietnam in 1967-68 and the extremely heavy exposures experienced in industrial accidents are not typical of the exposures of ground troops in Vietnam. Th« Veterans Administration has been evaluating Vietnam veterans for signs and symptoms of adverse health effects attributable to phenoxyherbicide and dioxin exposure. Data collection began as the "Agent Orange Registry" in 1978. Any veteran who was concerned about the health effects of Agent Orange could report to a V.A. hospital for a complete medical and exposure history, physical examination, and selected laboratory tests. By September J983 over 110,000 veterans had been evaluated and summary results had l«:fii published by the V.A. Although the quality of the evaluation received by t:hJ.H extremely large group of veterans is not in question, the data from tl>* VA Agent Orange registry are not suitable for CDC's epidemiologic study because the sample was self-selected. In order for CDC's study to be valid, a random sample must be evaluated to assure that results are not affected by selection bias. 6« Small Business Dat:-T collection will involve only individual subjects and controls selected according to the sampling procedure described in section B of this justification; no portion of the collection effort will involve small businesses or similar entities. 7. Consequence of Less Frequent Collection The data collection proposed herein is a one-time per subject effort; follow-up of these study and control groups is expected to include medical examinations of a randomly selected subset of subjects and periodic ascertainment of vital status of respondents. Vital status determinations can probably be accomplished by means of existing records systems. 8. 5 CFR 1320.6 It will be necessary to compensate the participants in the medical examination component of the study for their time, if we are to achieve a participation rate high enough to produce valid results. The examination itself will require at least two days, and way well extend into a third. Iravel to and from the examination site will require one day each way; thus the average time commitment per subject for examination will be 4-5 days. The length of time involved makes it impossible to schedule examinations entirely on weekends or other routine "off days," and many subjects will have employment which does not provide paid time off for �purposes such as this. Further, the number of subjects involved (10,OCO) makes it impractical for the Government to try to make individualized leave arrangements for each participant. The only option available to prevent the medical examination from being a prohibitive financial burden on the less well-to-do participants (thus producing a biased sample) is to pay a stipend to each participant. The Air Force in its Ranch Hand II Study compensated its participants at $100/examination day, and succeeded in attaining a participation rate of 95-f%. Participation at a similar level is highly desirable, especially when one is trying to detect rare events (e.g. certain forms of cancer). If CDC is to hope to approach the participation rate achieved in the Air Force study, it is clear that some similar compensatory arrangement will be required. 9. Consultation In developing this submission, CDC has had a number of "outside" consultations. These have included scientific reviews and contacts with other interested parties, principally veterans groups. In May, 1983, scientific reviewers were sent copies of the study protocols and invited to comment as were representatives of several veterans groups. In addition to the protocol review, CDC has conducted update briefings with veterans' representatives. The last such briefing was on August 31, 1983. The following is a list of scientists and veterans' representatives with whom CDC has worked. a) Scientific and/or Government Reviewers: 1. Agent Orange Working Group Science Panel 2. Howard W. Ory, M.D. Deputy Director for Research, EPO, CDC 3. Richard Dicker, M.D. Medical Epidemiologist, EPO, CDC 4. Dave Culver, Ph.D. Hospital Infections Program, CID, CDC 5. Claire Broome, M.D. Chief, Respiratory & Special Pathogen Branch, CID, CDC 6. Richard Remington, Ph.D., Chairman Vice President for Academic Affairs University of Iowa Iowa City, IA 52240 7. Margit Bleecker, M.D. The Johns Hopkins Medical Institutes School of Hygiene and Public Health Division of Occupational Medicine 615 North Wolfe Street Baltimore, MD 21205 �8. George L. Carlo, Ph.D. Epidemiology, Health & Environmental Sciences 1803 Building Dow Chemical U.S.A. Midland, MI 48640 9. Weal Castagnoli, Jr., Ph.D. Department of Chemistry & Pharmaceutical Chemistry University of California San Francisco, CA 94143 10. Theodore Colton, Ph.D. Boston University School of Public health 800 East Concord Street Boston, MA 02118 11. Mr. Frederic Halbert 12150 Banfield road Delton, MI 49046 12. George B. Hutchison, M.D. Harvard University School of Public Health 677 Huntington Avenue Boston, MA 02115 1.1. Patricia King, Georgetown Law 600 New Jersey Washington, DC Esq. Center Avenue, N.W. 2000.1 .14. Lewis Kuller, M.D. Dept. of Epidemiology Graduate School of Public Health University of Pittsburgh 130 DeSoto Street Pittsburgh, PA 15261 15. Claire 0. Leonard, M.D. 1445 Wilton Way Salt Lake City, UT 84108 16. John F. Soromer, Jr. The American Legion 1608 K Street, N.W. Washington, D.C. 2C006 17. Mr. Theodore P. Sypko Veterans of Foreign Wars of the United States V.F.W. Memorial Building 200 Maryland Avenue, N.E. Washington, C.D. 20002 18. Mr. John F. Terzano Vietnam Veterans of America 329 Eighth Street, N.E. Washington, D.C. 20002 �19. Mr. Monte C. Throdahl Sr. Vice President, Environmental Policy Staff Monsanto Company 800 N. Lindbergh Blvd. St. Louis, MO 63166 20. II. Michael D. Utidjian, M.D. Corporate Medical Director American Cyanamid Company Wayne, NJ 07470 21. G. Comstock, M.D. Johns Hopkins Medical Institutes 22. R. Hoover, M.D. NCI 23. R. Monson, M.D. Harvard University 24. J. Moore, M.D. NIEHS 25. P. Sartwell, M.D. Formerly of Johns Hopkins Medical Institutes 26. I. J. Selikoff, M.D. Mt. Sinai Hospital (K.Y.) �b) Veterans' Representatives 1. Mr. John Somraer The American Legion 2. Mr. John Terzano Vietnam Veterans of America 3. Mr. Fred Juarbe Veterans of Foreign Wars 4. Mr. Charlie Thompson Disabled American Veterans 5. Lewis Milford, Esq. National Veterans Law Center 6. Mr. Fred Mullen Paralyzed Veterans of America 7. Mr. Noel Woosley Am Vets 8. Mr. Jack P. Carver American Red Cross 9. Mr. Wilburn Long Blind Veterans of America 10. Mr. Frank Weil American Veterans Committee 11. Mr. Dick Gallant Military Order of tha Purple Heart 12. Mr. Dick Johnson Non-Commissioned Officers' Association 13. Mr. Max Beilke National Association for Uniformed Services Comments, recommendations, and criticisms received from reviewers are addressed in the final version of the study protocol. �JO 10. Confidentiality Assurance The Acting Director, National Center for Health Statistics, delegated to the Director, CDC, the following authorities under Title III of the Public Health Service Act, as amended, as they pertain to the epidemiologic and statistical responsibilities assigned to CDC. Section 304 of the Public Health Service Act (42 U.S.C. 242b), as amended - General Authority Respecting Research, Evaluations, and Demonstrations in Health Statistics, Health Services and Health Care Technology to collect information through health statistical or epidemiological activities, where such activities of CDC are not duplicative of other activities of the Department, and when the Director, CDC, determines that the authority to give assurances of confidentiality based upon Section 308(d) is necessary for the successful conduct of these statistical and epidemiological activities. Section 306 of the Public Health Service Act (42 U.S.C. 242k), as amended - National Center for Health Statistics, to collect information through health statistical or epidemiological activities, where such activities of CDC are not duplicative of other activities of the Department, and when the Director, CDC determines that the authority to give assurances of confidentiality based upon Section 308(d) is necessary for the successful conduct of these statistical and epidemiological activities. Section 308(d) allows an assurance of confidentiality to be authorized for the protection of identifiable information about individuals or establishments. Approval to give study participants assurance of confidentiality (Attachment 4, Confidentiality Assurances Statement) under these authorities has been requested from the Director, CDC. Verbal approval to assure confidentiality has been given; a copy of the formal authorization will be forwarded on receipt. 11. Sensitive Data Much of the data to be collected in these studies can be considered sensitive. Questions will be asked regarding race, religion, legal difficulties, employment problems, fertility problems, and illicit drug use. Race and religion information must be collected, because some conditions of interest (e.g. cancer) are not randomly distributed with regard to these factors. Questions about legal difficulties, employment problems and illicit drug use are necessary because veterans groups have suggested that these conditions are in excess among Vietnam veterans; that contention must be evaluated. Finally, information about fertility problems is required because increased rates of infertility and birth defects have been attributed to Agent Orange exposure. �11 12. Cost to the Federal Government Conduct of these studies will involve both "in-house" and contract expenses in excess of $73,000,000 over a period of four years. Costs will be borne by the Veterans Administration, and outlays are projected in the following amounts* for the categories shown. Object Class 1984 1985 1986 1987 1. Personnel $3,000 3,150 3,300 3,040 12,490 2. Travel/Transport of Persons Employee Travel All Other Total 300 60 315 20 330 15 275 10 1,220 105 3. Trans, (things) 30 30 10 10 80 4. Cotnmo/Utilities (& other rent) 50 60 60 50 220 5. Printinp & Repro. 25 50 50 30 155 10,576 19,320 19,320 9,660 58,876 7. Supplies & Mtls. 10 10 10 10 40 8. Equipment 25 15 15 15 70 14,076 22,970 23,110 13,100 73,256 6. Contracts Totnl *In Thousands Ko dJrect costs will accrue to the study participants. Interviews vill be scheduled at times that do not conflict with the particular respondent's work, and participants in the medical examination component of the study will have no out of pocket expenses for travel, lodging, subsistence, or incidentals associated with the examination.. The examination itself, of course, will be free to the participants. �12 13. Respondent Burden The Agent Orange and Vietnam Experience studies will involve 30,000 respondents (5 cohorts, 6000 subjects per cohort). It is anticipated that 85% of individuals falling into the sample will be locatable and that 85% of those people will agree to interview. Thus, a sample of 8350 subjects will be drawn of whom 7100 should be locatable and 6000 of those interviewable. Average contact time is expected to be about 55 minutes. Contact with refusals will be brief ( JO minutes) while complete interviews may require one hour or more ^ depending upon the extent and complexity of responses. Interviews will be conducted by telephone on a one time per respondent basis. At least two thousand subjects per cohort will be selected randomly and asked to participate in a thorough medical and laboratory evaluation. Individuals falling into this subset will be contacted a second time to secure their participation in the examination phase of these studies. Burden hours for the Agent Orange and Vietnam Experience are projected as follows: 1984 Hour.-; 6,000 1985 1986 1987 10,000 10,000 4,000 The .Selected Cancers Study will involve approximately 1300 cases and 1300 controls; average interview time for both cases and controls will be one hour. However, the fatality rate for soft tissue sarcoma is quite high, and it may be necessary in some cases to collect information from next-of-kin instead of tbv nffected man. In these situations data collection would be limited to relatively simple items such as whether the iran served in Vietnam. Thus, next-oJ-kin interviews will be extrerrely brief. Sinre the cases of interest are those occurring from July 1, 1984, to June 30, 1988, respondent burden, for both cases and controls, will be spread over four years. Interviews will be conducted by telephone on a one-time per respondent basis; distribution of burden hours is expected to be as follows: Total burden hours for all three studies are: 1984 Hours 1985 1986 6,650 10,650 10,650 14. Changes in Burden At this time there is no cause to expect changes in the estimate of respondent burden. Should field experience suggest an increase or decrease, an amended estimate will be submitted. �13 15. Project Schedule CDC will prepare comprehensive reports of the findings for each of the study phases; ideally, major findings will be published simultaneously in peer-reviewed medical journals. Contingent upon necessary funds and positions being available, the following timetable is proposed for the three study phases: 1. Way 84 - August 84: Sample selection & pilot testing for Agent Orange (AO), Vietnam Experience (VE), and Selected Cancers (SC) studies. 2. October 84 - January 86: AO, VE, and SC main study interviews and exams. January 86 - March 87 4. March 87 - September 87; B. Complete AO, and VE main study interviews, exams, and mortality data collection. Report findings. Report SC study data. Collection of Information Employing Statistical Methods * • Respondent Universe Th<; potential respondent universe for the Agent Orange Study includes all non-officer single term enlistees and draftees who served in the Array in III Corps in Vietnam in 1S67-68. The Vietnam Experience universe Is all non-officer tingle term enlistees and draftees who served in the Army during the period 1966-71. The universe for the Selected Cancers Study is all men vith 1S29-1'J.'.>3 birthdates who reside in 10 or more SEER areas (NCI, 1981). The probable areas are: The states of Connecticut, Hawaii, Iowa, New Mexico, Utah, and the Commonwealth of Puerto Rico; and the metropolitan areas of Atlanta, Detroit, San Francisco and Seattle. Numbers of potential respondents, sample sizes, and participation rates are displayed in tables I-V below. �14 I. Agent Orange Study, Interview Phase Universe - 200,037* (individuals; 1st Term Army, Non-Officer, 1967-1968) Full Sample Locatable (? 85%) ( Interviewed ( ? 85%) ( Cohort #1 8350 7097 6032 Cohort #2 8350 7097 6032 Cohort #3 8350 7097 6032 Cohort #2 2410 2410 2000 Cohort it3 2410 2410 2000 II. Agent Orange Study, Clinical Phase Full Sample ( ) # Locatable (@ 100%) Examined ( ? 83%) ( III. Cohort 2410 2410 2000 Vietnam Experience Study, Interview Phase Full Sample (//) Low table (£< !;.$%) Interviewed ((; 85%) Cohort #1 8350 7097 6032 Cohort #2 8350 7097 6032 IV., Vietnam Experience Study, Clinical Phase Universe = 12,064 (interviewed subjects) Fu.ll Sample (ft) Locatable (G'100%) Examined (@ 83%) Cohort ifl 2410 2410 2000 Cohort #2 2410 2410 2000 Cases 1228** Controls 1800 V. Selected Cancers Study Universe » 2,481,000 Total (4 years) * Value was derived by applying the percent of "all Army" assigned to Vietnam in 1967 (18%) and 1968 (21%) to the total number of inductions and first enlistments in those years; 489,389 in 1967 and 533,082 in 1968. Intuitively one would expect "first termers" to be more often assigned to Vietnam than their percentage of "all Army" would indicate; however, no data are available with which this supposition can be evaluated. ** Estimate based on rates of sarcoma, lyinphoma, nasopharyngeal, and liver cancers in selected SKER areas. �15 2. Data Collection Procedures CDC proposes to limit this study to draftees and single term enlistees in the non-officer ranks who served in the Army; selection will be further limited to those who had only one tour of duty in Vietnam. Exclusion of officers is based primarily on a desire to make the groups as homogeneous as possible with respect to pre-existing demographic factors which could influence health. In addition, the inclusion of officers might require substantially increased record review to assess herbicide exposure potential (see below) because of multiple tours of duty in Vietnam. Exclusive focus on veterans of the Army is chosen for several reasons. Ihe Army had a much greater proportion of draftees than the other services and it is felt that it is important to include substantial numbers of them in the study. Use of draftees will probably make achieving a balance on such factors as training, military occupational specialities, and pre-existing demographic factors easier. Inclusion of substantial numbers of draftees is also motivated by a desire to try to make an assessment of the possible association between volunteerisai and health. (However,such an assessment may not be possible if a large percentage of enlistees joined the Army because they felt that the draft was inevitable.) CDC proposes to exclude the Marine Corps in part because its men were mostly volunteers and in part to limit the amount of records review required to select study subjects (the reasons for this will be better appreciated after the selection process is described). In addition, the AA01F has? worked most extensively with the records of the US Army, has ' become cost familiar with them, and feels most confident about their quality. Moreover, the Air Force did not keep records which allow the daily geographical placement of personnel, and there were rather limited numbers of Navy serviceri«n who were stationed on land in the Vietnam theatre. Even though all sf.udy participants will be males in the non-officer ranks who were in the Aoiy, It is likely that the results will be useful in making inferences about all men who had similar ground experiences and possible herbicide exposures in Vietnam; the same may be said about females if there are no sex-specific effects. As has been noted previously, there vill be three cohorts of men chosen for the Agent Orange study. The first two cohorts, which will differ with respect to the likelihood of exposure to herbicides, will be chosen from III Corps (en area where herbicides were used extensively) during the same period of titie, 1967-1968. This will be done in order to make the two as similar as possible with regard to the nature of their service 'experience — similar with regard to, for example, type of terrain, Indigenous diseases, and intensity of coabat. To enhance the possiblity of including soldiers who may have been exposed to herbicides, the men Included in these first two cohorts will ba selected exclusively from combat battalions. Since these two cohorts will be chosen from an area where herbicides were extensively used, there Is potential for exposure misclassificatlon (i.e. some of the supposedly unexposed veterans may In fact have been In contact with herbicide). The third cohort will therefore be chosen from an area where there is good evidence that there was no usage of herbicides. According to the staff of the AAOTF it will probably not be possible to derive this third cohort exclusively from combat battalions. �16 Selection of veterans to be included in the first two Agent Orange study cohorts will be clone by a multi-step review of military records, beginning with the selection of a geographical area of consideration and ending with the choice of individual soldiers. Since many of the proposed procedures are untested, modification may be required after pilot study assessments . In suamary, the steps required are: 1) 2) 3) 4) 5) 6) 7) select a geographical area and time of interest - these will be III Corps and 1967-1968 determine which of the battalions stationed in III Corps in 1967-1968 have acceptable records choose a random sample of 50 battalions (250 companies) from among all battalions with acceptable records abstract selected companies' locations on one randomly selected day of the week for each of the 104 weeks in 1967-1968 using the "Herbs" and "Services Herbs" tapes, score the herbicide encounters of the 250 companies on the 104 days rank the 250 companies with respect to their herbicide encounters choose men for the "likely exposed" cohort from companies at the top of the ranked list and men for the "likely not exposed" cohort from tho.se at the bottom of the list. The rationale for these steps is presented below. In order to limit the amount of records review required, the first step is to restrict, on the advice of the AAOTF, the geographical area of consideration Co III Corps and the time period to 1967-1968. This time period and area was selected because of a variety of factors, including the number of P.anch Hand missions, the relatively high level of TCDD contamination of the Agent Oranj't- used then, and U.S. troop strength, which was at its peak. The AAOTF has determined that there were about 110-120 Army combat battalions stationed iu III Corps during that time (usual battalion strength was 1000). The records of the companies attached to these battalions will serve as the major source of information about troop locations. The second step in the selection process will consist of a review of GSA documents to tiscertain which battalion records appear to have unacceptable time gaps (if gaps appear in battalion records it may be possible to supplement them vith division and brigade level records, and this will be done when feasible). CDC does not feel that It is necessarily wise to exclude a unit simply because some of its records are missing — units with missing records could have had more or less exposure to herbicides than units with complete records. Therefore it is proposed to apply the following criteria regarding records quality: if a battalion has more than 30 contiguous days of absent records or an aggregate of more than 60 days absent records for the time period 1967-1968, the unit will be considered unsuitable for inclusion in the study. If very few units are found to have gaps of this magnitude it is possible that more stringent criteria can be used. For each of the combat battalions located in III Corps in 1967-1968, the AAOTF will summarize the condition of the records as indicated in the GSA documents. �17 The third step will be the choice of a random sample; of 50 battalions (250 companies) from among those which are judged suitable during the second step. Step four will involve abstracting from company records (or battalion records, if necessary) all locations recorded for the selected companies on each day for each of the 104 weeks during 1967-1968. These two sampling steps will be done in order to limit the quantity of records review required, but it should be sufficient to provide a reasonable estimation of the range of herbicide encounters. CDC believes that this is an important issue — at this point the frequency and nature of troop herbicide encounters is largely a matter of conjecture (aside from the work done by the AAOTF with 2 Army battalions). As noted before, the records available will never permit an unambiguous assessment of exposures, but this approach will help to place a frame of objectivity around the issue, at least for men in Army combat units in III Corps in 1967-1968. In step five, CDC will check the selected company locations against the locations of herbicide applications as recorded on the "Herbs" and "Services Kerbs" tapes. The "Herbs" tape contains computerized records of Ranch Hand missions (time, place, type and amount of herbicide). The National Academy of Science report (1974) on the effects of herbicide usage in Vietnam contains a relatively .limited assessment of the accuracy of these records. CDC finds the results of this Investigation encouraging, but doubt about accuracy exists in some quarter:? today. CDC has requested that the National Academy make available tin' results of other checks which were done at the time, and to look into the possibility of further accuracy checks. The "Services Herbs" tapa primarily contains records of non-Ranch Hand herbicide applications (eg, base perimeter sprayings). This set of data has been put together by the AAOTF froE a review of a variety of military records; the degree of completeness of the "Servicf?K Herbs" data set is unknown. The number of unit encounters with herbicide applications according to these data t;«-tt; will be tabulated by at least three systems. The first of these systfliis will have geometrically progressing scores or weights for various space and time distances and the second will have linear weights. The aggregate scores for these two systems will be based on the products of the tire and spar.p scores. The third system, a variant of one proposed by the Department o.l Defense, will simply count the number of encounters which are at distances of l^ss than 3 days and 2 kilometers. The purpose of these exposure systems is to obtain a spread of unit exposures so that units can be chosen from the top und bottom of the scales. It is desired that the spreads obtained should reflect "meaningful" differences in exposure. Relatively little is known about the environmental fate of herbicides and TCDD, and even less is known about the human pharmacokinetics of these substances. Because of this lack of knowledge, these systems are necessarily arbitrary and this motivates the proposal of three scales. The scorings for the first two systems proposed for preliminary tabulation ara indicated below. �18 Exposure System A 1. Ranch Hand Missions a. Regular Missions — cross-classified by time after mission «=! day, score=16; 2-3 days, score=4; 4-30 days, score=2; and 31-59 days, score^l), distance «=1 km, score=4; 2-3 km, score=2; 4-8 km, score=l), and type of herbicide. b. 2. Aborted Missions — cross-classified and scored as above. Other Herbicide Applications (e.g., perimeter spraying)—for those encounters <= 1 km classified by tiae and scored as above Exposure Sy«t«m U. 1. Ranch Hand Missions a. Regular Missions — cross-classified by tine after mission «»1 day, score=4; 2 - 3 days, scoraa3; 4 - 3 0 days, score=2; and 31 - 59 days, score=l), distance «=1 km, score=3; 2 - 3 km, sc.ore=2; 4 - 8 ka,score=l), and type of herbicide. b. Aborted Missions — cross-classified and scored as above. 2. Other Herbicide Applications (e.g., perimeter spraying) — for those encounters <« 1 km classified by tine and scored as above. �19 As mentioned before, the various encounters will be weighted by the product of the time and distance scores; each encounter of a unit with a particular herbicide application will be counted in only one time and one distance category. For example, using Exposure System A an encounter with a Ranch Hand mission within 1 day and 1 km would receive a score of 64, as would an encounter with a base perimeter application within 1 day (small bases); an encounter with a Ranch Hand application within 4 - 3 0 days and 2 - 3 kilometers would get a score of 4. Using the third (modified Department of Defense) system, any encounter which occurs within the 3 day-2 kilometer limit vould receive a score of 1. The daily scores determined by each of the three exposure systems will then be summed over the sampled 104 days for each company. Kext, the 250 or so companies will be ranked on their summed encounter scores. If there is good agreement in the rankings provided by the three systems, those at the top of the lists will provide individuals for the "more exposed" cohort and those at the bottom will contribute to the "less exposed" group. If there are substantial disparities in the rankings provided by the three systems then roughly 1/3 of each of the two cohorts will be chosen from the top and bottom of each of the rankings. At this time it is unclear how cany companies will have to be selected to provide the requisite number of individuals for these 2 cohorts, but it will probably be on the order of 50 to 60 from the top and a like number from the bottom. If 55 companies each provide 150 suitable individuals this number will allow some loss due to non-participation and yield the number desired for each of the cohorts. The desire to omit the Marine Corps from this study can now be more easily understood. If Marines were to be included, the records review and other selection tasks to this point would have to be done eeparately for them because they were largely stationed in I Corps, and this would cause delay. The next stop will be the choice of individual soldiers from the selected units. This process will begin with a review of company morning reports. Individuals who appear to meet the criteria with respect to type of entry into the service (draftee or single terra enlistee), are in the non-officer ranks, and whose 1-year Vietnam tour began and ended during 1967-1968 will be considered potentially eligible for inclusion in one of the cohorts. For those vho appear to be eligible, the AAOTF will also document their presence or absence with the selected units on each of the days during the 2 year period 1S67-1968. Those individuals who were absent from their units for more than 90 days of their scheduled 12 month tours (exclusive of their regular R&R lo&ve) will be considered ineligible for final selection. The AAOTF will also document the reasons for all absences for both the selected men and those men who would be eligible save for their absences* Thus, this process will provide CDC with, inter alia, a measure of combat intensity since absences for reason of casualty will be recorded. Individual personnel folders will be obtained from the St. Louis records center by the AAOTF for soldiers considered eligible. Staff of the AAOTF will abstract certain identifying and service (e.g., military occupational specialty) information from the individual personnel folders and forward the information to CDC on an incremental basis so that it can begin the process of locating the veterans and soliciting their parlcipation in the studies. Company records will also be used to document the locations of the selected units on all days during 1S67-1968. This information will later be used to classify individual soldiers with respect to exposure to herbicides by a scheme similar to that noted above. �20 The third cohort for the Agent Orange study will be selected by a different method. Areas in Vietnam where there is no evidence of herbicide usage prior to 1969 will be identified by the AAOTF and a roster of units which served in, and only in, those areas and only in those areas in 1967-1968 compiled. The staff of the AAOTF has suggested that Cam Ranh Bay or Vung Tau might be examples of such areas. Enough units will be randomly chosen from this roster so that the required number of individuals can be included in the study. The eligibility criteria for selecting individuals from within the selected units will be the same as those used for the first two cohorts. The AAOTF will provide CDC with the same sort of identifying, service, and absence information as it provides for those individuals included in the 2 other cohorts. Vietnam Experience Study The procedures for selecting individuals for the Vietnam Experience study will be substantially different from those used for the Agent Orange study — the process will start with the selection of individual personnel files in the National Personnel Records Center in St. Louis rather than with the selection of military units. We understand that, for draftees and single term enlistees in the Army infantry, assignment to Vietnam or to some other part of the world was essentially a random process, but this was probably not the case for other services. Since it is desired to compare men who went to Vietnam with men who did not, but who had a more or less equal chance of being assigned to Vietnam, CDC will limit this study to Army veterans in the non-officer r.nnks. The St. Louis records center houses personnel files for all discharged service persons, except the living retired end those who are in the active reserves. Soon after discharge, the military personnel folder is transmitted to the center where it is identified by service and given an accession number. Since a n>.i:i(:yr list by service and accession number is available it is possible to select a sample of individuals from the records center stacks. Unfortunately, the master accession list does not indicate whether the discharged ooldier served in Vietnam or not, nor his rank, nor any other vitalinformation. Thus it will be necessary to pull the records of each individual identified from the accession list to determine if he qualifies for inclusion In the study. Those individuals found to be ineligible will be replaced with another serviceman according to strict criteria. This eligibility assessment will be done at the records center, and coordinated by AAOTF staff; records of individuals found to be eligible at this preliminary review will be sent to AAOTF headquarters in Washington, D.C. for complete review. CDC and AAOTF staff visited the St. Louis Records Center and reviewed a random sample of 1259 military records. Of this sample, 563 records were of veterans who met the preliminary study criteria for inclusion. Of those qualified, 43% had served in Vietnam, 21% in Germany, 7% in Korea, and almost all of the remaining 29% served only in the United States. The distribution by location of service and time of that service correspondends to Department of Defense data. This work indicates that the approach can yield a sample with relatively little wasted effort and CDC feels that it is far preferable to a sampling scheme based on a preliminary selection of military units. The members of both cohorts for the Vietnam Experience study will be chosen from among soldiers with appropriate periods of active service. For the Vietnam service cohort this should provide a year-of-tour distribution which is proportional to the year by year Army troop strength in Vietnam over the period 1966-1971. The selection procedure for the control cohort will be �21 such that Its period of service distribution is equivalent to the Vietnam cohort. The cohort of men Included in the Vietnam service cohort will have served only in the U.S. and Vietnam. It is proposed that the control or non-Vietnam cohort be chosen so that it comprises 3 groups: a group of m»n who served only in the continental US, a group whose members served in the U.S. and Europe and a group of those who served in the U.S. and Korea. This approach may allow an assessment of the effects of the experience of a foreign service, with the contrast between European and Korean service providing a contrast in the level of foreign environment of the duty stations. AAOTF will give CDC the came sort of information about each soldier In this study os will be provided for those men included in the Agent Orange study, except that no daily geographical location information will be given. Data collection will be identical in the Agent Orange and Vietnam Experience studies; it will entail telephone interview of each locatable cember of each study sample. (Attachment 5, AO/VE Questionnaire). Interviews will be performed by a competitively selected contractor who will be responsible for developing all supplementary forms, letters, etc. and for generating whatever additional locating information is necessary to complete the required interviews. CDC will provide the contractor with a monthly list of approximately J400 potential participants; government supplied Information on each subject will include: name, date of birth, SSAN, last known address, and the name(s) and address(es) of next of kin (extracted from military records). The contractor will be required to verify the addresses provided or develop new ones and to determine the subjects' telephone numbers. Initial contact with each subject will be by JU-tter, and the contractor will be required to exhaust all locator systems hefor» contacting next of kin to establish subjects' whereabouts. CDC tesf>-il the "locatability" of veterans of battalions using the IRS record system and telephone directory assistance. The Army Agent Orange Task Force identifcled 840 veterans, and IRS records match was made on 754 (89.8%) of them. Directory assistance verified address and provided a telephone number for 360 (47.9%) of those individuals; verified address, but provided no telephone number (unlisted)for 106 (14.1%) more. Directory assistance was unable to match name and address for 286 (38.0%) subjects; however, in 36 of thsse "no match" cases the operator indicated that there was a listing for the name at a different address; so the chances of locating those individuals should bfi quite good. Based on this experience, the contractor can reasonably c:':pftct to be able to locate approximately 67% of the subjects provided simply by using government provided information and by contacting directory assistance. The first mailing to a potential respondent will identify the study and explain its purpose; it will also contain a toll free (800) number which a potential respondent can call for additional information. (Attachment 6, Initial Contact Letter—Draft). The contractor will be responsible for developing the final contact letter. This mailing will indicate the voluntary nature of participation and will also inform the subject that he will be contacted by telephone for an Interview. First telephone contact will be made by an interviewer who will explain the purpose and procedures of the study, its. voluntary nature, and attempt to elicit agreement to participate (Attachment 7, Draft text, initial telephone contact). If the subject is willing to proceed, the interviewer will administer the questionnaire. If a practical method for doing so can be devised, the interviewer will be blinded to the cohort status of the respondent. �22 Subjects who do not respond to the initial mailing or to whom the first letter is undeliverable will be included in a locating system which involves at a tuniiaum, telephone company sources, credit bureau, post office .forwarding, and DMV (driver's license). The contractors final locating effort will be to contact the subjects' recorded next(s) of kin. Contact with next of kin would be by mail with possible telephone follow-up. Subjects who initially decline to participate will be contacted three tiires before being finally classified as refusals. Two attempts to motivate participation will be made by telephone and a final effort in person. If the field worker is unsuccessful in eliciting cooperation, he will attempt to ascertain the subjects ' reasons for non-participation and terminate the contact. At no time will study representatives use coercive methods to secure subject cooperation. Selected Cancers Study As noted before, this part of CDC's efforts to address concerns of Vietnam veterans will take the form of a population-based case-control study. A case-control study will be conducted because a cohort study would require truly massive sample sizes to detect an increased risk for such rare diseases, much larger samples than those proposed for the Agent Orange and Vietnam Experience studies. Studying such large samples would unnecessarily delay CDC's ability to provide answers to veterans about their risks for more common disorders. The term population-based implies that all cases of sarcoma, lymphoraa, nasopharynxes1, and liver cancer in defined population groups will be ascertained and an attempt made to include them in the study. This will confer at len.st two major advantages over studies done with cases collected by other methods: 1) since all cases arising in a population are ascertained, the concerns about biases of ascertainment which always attend other case selection strategies are not at issue, and, 2) a population-based study allows estimates of attributable risk, not just relative risk. The control group will be chosen from the same population as is the case group, and this xd.ll allow estimation of disease incidence rates by veteran status. It is proposed to use the Surveillance, Epidemiology and End Results (SEER) Centers, which are sponsored by the National Career Institute, as the source of cases. The SEER Centers ascertain nearly all people newly diagnosed with cancer in at least 10 defined population areas (National Cancer Institute, 1981). These areas are: the states of Connecticut, Hawaii, Iowa, I.'ew Mexico, Utah, and the Commonwealth of Puerto Rico; and the metropolitan areas of Atlanta, Detroit, San Francisco, and Seattle. All of the SEER Centers contacted by CDC have indicated that they are interested in participating. Overall, interest in participation appears high because the SEER centers want to continue to build and demonstrate their epidemiologic potential. In addition, the centers each employ at least one epidemiologist, many of whom have been involved with the issue of cancer and chemical exposures and who view the proposed study as personally interesting. Overall, CDC believes that the SEER network is a superb epideniologic resource that has been proven in other large case-control studies such as those which investigated the association of bladder cancer with artificial sweetener use (Hoover et al., 1981) and uterine, ovarian, .and breast cancer with oral contraceptive use (Layde et al., 1983). Other population-based cancer registries may be utilized for case ascertainment if they are Interested in collaborating in this study and if their case ascertainment is complete and rapid enough. �23 All cases of coft tissue sarcoma, lymphoma, nasopharyngeal and liver cancer occurring from July 1, 1984, to June 30, 1988, in males with birthdates 1922-1953 who reside in the geographic areas covered by the participating population-based cancer registries will be included in this study; the cases will be contacted and interviewed within 3 months of diagnosis. This age group has been selected because it includes the men most likely to have served in Vietnam between 1965 and 1971. Since soft tissue sarcomas are so rare, CDC has considered including additional cases diagnosed prior to July 1, 1984, in order to increase the power of the study to detect an association which way be present between herbicides and/or service in Vietnam and sarcomas. This possibility has been (tentatively) rejected for two reasons: 1) most importantly, the Swedish studies which suggest a relationship between sarcomas and occupational exposure to 2,4,5-T indicate a mean latency period between first exposure and diagnosis of about 16 years. Therefore, including cases which arose prior to 1984 might give only an illusion of Increased power; 2) because the fatality rate for soft tissue sarcoma is quite high (Tucker et al., 1982), information about early cases and controls would frequently have to be gathered from next-of-kin instead of the affected man. However, this latter point would not be a major concern if data collection for these cases was limited to relatively simple items, such as whether the man served in Vietnam. four hi/a o'logic review panels each composed of 2-3 pathologists will be established—one group to review each type of cancer. The groups will receive a set of slides or tissue block on each case and will establish their own diagnosis wjtlunit knowledge of the presumed diagnosis. Interviews with cases will not b* tU-J.iyed for confirmation by the pathologic review panels. The sel.M-tiou of controls will be by the method of random digit dialing (HDD). Te.l?:-phone numbers are randomly phoned and a brief census of the household ifi made. If a man of the right age is found, then he will be asked to participate in the study. This method worked successfully in the National Cancer Institute Bladder Cancer study (Hoover et al., 1S81) and CDC's Cancer and Steroid Hormone Studies (Layde et al., 1983). Over 90% of households that had eligible won>en in CDC's study yielded an interview; the NCI results were similar. Unlike the usual methods of collecting a sample of a population, v?hich depend on making at least a partial in-person census of the geographic area, RJ)D allows this to be done by telephone, which clearly is less expensive and far more practical. About 95% of households have telephones. In addition, several researchers have documented how well samples chosen by RDD reflect the general population. The main concern is that people of very lov -socio-economic status may be underrepresented in the control group. CDC feels the effect of this potential bias will be small for 2 reasons: 1) our control group vill be so large that some very poor people will be included; 2) an analysis stratified by socio-economic status should help ameliorate whatever bias is present. Based on the age and race distributions of cases, CDC will select controls from the list of eligible men such that the overall age and race distribution of the controls will be similar to that of the cases. As the study progresses, if the age distribution of cases is different from expected, control selection can be modified. Data collection for the Selected Cancers Study will differ from that in the cohort studies previously described in that a different questionnaire will be used and it will not be practical to employ a computer assisted telephone interview due to the relatively small number of subjects available for interview at any one time. �24 The SEER Centers which identify the coses of interest will also perform the interviews (Attachment 8, SCS Questionnaire). CDC will select controls by ir,eans of a random digit dialing process.. Potential controls will be informed by telephone of the purpose of the study and its voluntary nature; they will then be asked if they would be willing to participate in a telephone interview if selected. Those individuals who agree will be asked for age and race information (for matching purposes) and included in the pool of potential controls. Contact and interview procedures will be the same for both cases and controls. The participating SEER Centers will send a letter explaining tha study and its voluntary nature to each case/control (documents to be developed by contractors). Three days after the initial mailing an interviewer will make a follow-up telephone call to answer questions and make an appointment to complete the telephone interview; ideally, the interview will not be aware of the case/control status of the subject. Experience in similar studies suggests that participation rates will be relatively high (ca 90% of cases; at least 75% of controls); thus no elaborate motivating procedure has been established. Both subjects and controls who are undecided or who initially refuse interview will be called a second time by an interview supervisor who will attempt to secure participation or, at least, ascertain the reason(s) for refusal. Sample Sizes, Statistical Power and Participation^Rates Agent Orfitige^ and Vietnam Experience Studies j. Reasonse_ Rates and Power The sensitivity (power) of these studies to detect a real increased risk among the veterans in any one of the cohorts depends on several factors, most prominently the numbers in each of the cohorts, the prevalence or incidence of the condition of concern, the amount of misclassification on the variables used to define the cohorts, and the magnitude of the increased risk. It is proposed that each of the cohorts included in the mortality follow-up and health interview phases of these studies be composed of 6000 ir.en. The number 6000 was chosen since this will give good power (beta-alpha-0.05, 1 tail) to detect a 2-fold increase in the risk for health outcomes normally occurring at the rate of about 5 per 1000 in comparisons of two cohorts (if there is little or no misclassification In the selection of men for the cohorts). A high beta level, equal to the alpha level, is suggested since CDC believes that as much attention should be given in these studies to type II errors os to type I errors. CDC further recommends that a sample of 2000 be selected from each of the cohorts for the medical, psychological and laboratory phase of the studies. This number is suggested since it will provide good power (beta*alpha«0.05, 1 tail) to detect 2-fold increases in the relative risk for health outcomes which ordinarily occur at the rate of 1.5-2.0%. A major limitation of the sample size calculations for the cohort studies is that no good data exist on the expected prevalences of the outcomes postulated to be associated with TCDD exposure in populations similar to the veterans being studied. The occurrence of many of these conditions has never been assessed in population-based surveys. For some conditions there are data for men of the relevant ages from NCHS's Health Interview Survey (HIS) and Health and Nutrition Examination Survey (HANES). However, these national surveys may not accurately estimate the rate of chronic diseases in veterans — men who had to pass fairly rigorous medical examinations to get into the Army. In a sense, we will not be certain of the actual statistical power to detect increases in specific diseases until the analysis is underway and we know the frequency of the specific diseases in the unexposed cohorts. �25 Perhaps this discussion begs the question: How were the sample sizes for each cohort of 6,000 for mortality assessment and interview and 2,000 for examination and laboratory testing chosen? Because of the paucity of relevant prevalence data these choices were necessarily somewhat arbitrary, however, CDC believes they are appropriate to detect an increased risk of important health outcomes in exposed veterans. For example, the cumulative total cancer incidence in the "unexposed" groups of veterans from 1968 to the time of the interviews is expected to be about 6 per 1,000 based on data from the Surveillance, Epidemiology, and End Results (SEER) network of the National Cancer Institute. Therefore, we will be able to detect a 2-fold increased risk for this critical outcome (and all outcomes that occur in more than 5 per 1,000 of the unexposed). For the examination and laboratory testing phases we should be able to detect 2-fold increased risks of abnormal outcomes for dichotomous variables that occur in more than 1.5% - 2.0% of the unexposed. Based on HIS and HANES, these should include such important conditions as ischemic heart disease and diabetes mellitus. For continuous outcome variables, such as the results of roost laboratory tests, we should be able to detect even modest differences between the exposed and unexposed groups. The power calculations have been made on the assumption that categorical data analysis will be done on the basis of a single 2x2 table for each disease. It is very unlikely that the situation will be simple enough to allow such straightforward analysis. Rather, it is anticipated that analysis will involve multiple variables and this may reduce power, if unnecesary variables are inadvertently included. Although the reduction should not be great, the situation is far too complex to allow any a priori estimation of just how large it may be. Another factor which may reduce power is misclassificyt.ion of the variables used to define the cohorts ("exposure" variables) — Jf the mlsclassification is random. Of particular concern is the possibility that the records which have to be used to define the first two Agent Orange study cohorts ("likely exposed" and "likely not exposed") are so incomplete and/or inaccurate that there will be a sizeable amount of random wisclassification in respect to true herbicide exposure. If this is the case then power will be reduced, possibly to a significant degree, and the measures of effect will be biased toward the null. If misclassification in respect of exposure is present and not random, power would also be affected and the treasures of effect could be biased toward or away from the null. In order to achieve the power desired in the interview phase it will be necessary to b«.»}*in with cohorts which are larger than 6000 because some of the desired study participants will not be located and some, once located, will decline to participate. CDC recommends that the goal for this phase should be a location rate of 85% and a 85% interview rate among those located, for an overall participation rate of 72%. Therefore, CDC recommends that the AAOTF select 8350 (approximately 6000/0.72) veterans for each of the cohorts. If the interview phase is successful, it should not be difficult to obtain the cooperation of 2000 wen per cohort for the examination phase. However, there is considerable concern that we may have difficulty in achieving a high rate of participation among those who are selected for inclusion in this phase. In other words, our concern here is not that we will be unable to reach the desired sample size of 2000 per cohort but rather that participation is not limited to a highly selected group of men. It is felt that the best we can hope for is a rate of 60% cooperation (i.e., 83% of the subsample composed of those who are located and agree to be interviewed (0.83»0.60/0.72J). This may be an optimistic goal. The Ranch Hand study team had an examination phase participation of 87% among the Ranch Banders and 76% among the controls. CDC feels that the Air Force success can only be a goal which we can hope to �26 emulate but not necessarily achieve. The NCHS experience of about 70% participation in its Health and Nutrition Examination Surveys can also be considered (the. interview survey cooperation was about 95%). CDC feels that inferring directly from this experience to its own situation probably gives a somewhat optimistic expectation. The NCHS examinations were done in trailers which were located within easy commuting distance of the study participants, whereas most of CDC's study subjects will have to be transported to the examination sites by air. Moreover, the NCHS sample included persons of both sexes and all ages while CDCs cohorts will be composed wholly of men of a narrow age range, a group which will probably have a lower than average propensity to participate. It will be desirable to assess study participants and non-participants with respect to differences in health and differences in exposures to health-influencing factors. Soo.e assessment of this sort will be possible for the examination phase—men who are interviewed and who are invited but decline to participate in the exams will be compared to men who are examinsd. This comparison will make use of data gathered in the interviews. Unfortunately, a similar type of comparison cannot be made for those who are interviewed and those who are not. CDC will have very little,.if any, health related information about men who will not participate or who are not located. If feasible, comparisons will be made between interview respondents who readily participate and those who agree to be interviewed only after considerable coaxing. Similar comparisons could be made between veterans who are easy to locate and those traced only with considerable difficulty. While not ideal, such comparison# may provide insights into the characteristics of those refusing to participate and those not located. Selected Cancers Study As with the cohort studies, the power of this study to detect a real increased rlt;k among Vietnam veterans depends on several factors, in this instance the number of cases and controls interviewed, the proportion of controls who served in Vietnam (and/or the proportion exposed to herbicides), the amount of exposure tiisclassification (misclassification of disease should be held to a minimum through the use of panels of pathologists, and the magnitude of the increased risk. The Veterans Administration estimates that 2.9 million veterans served in Vietnam. As of July 1, 1S83, the United States civilian tnale population aged 30-55 was estimated to be 34,253,000. Therefore, it is estimated that 10 to 15% of males in the age group of Vietnam veterans (birthrates 1929-1953) actually served in Vietnam. Power figures for this study are presented in Table VI. We have decided to study about 1,300 controls since this number will give fairly good sensitivity for a 2-fold increase in risk, and adding further numbers to the control sample will do little in terms of improving the power. It is unlikely that small real increases in risk can be demonstrated. Moreover, if Agent Orange or some other factor really has increased the risk of exposed veterans a small amount, and If only a small porportion of veterans were exposed to a toxic dose, the sensitivity of this study will be much lower than the figures presented. It should be noted that this will be a large case-control study, based on all soft tissue sarcoma, lyrophoma, nasopharyngeal, and liver cancer cases which have occurred in a population of about 2,481.,000 males aged 30-55 over a period of 4 years. Viewed from a somewhat different perspective, it will have roughly the same sensitivity as a cohort study which assembled about 10% of all Vietnam veterans (290,000) and the same number of non-veterans and assessed the occurrence of soft tissue sarcomas over a period of 6 years and lymphomas over a period of 3 years. The cost of such a study would far exceed the cost of the proposed study. �Table Vi .1 Power1 of Selected Cancers Case-Control Study to Detect Increased Relative Risks a) 2~foJ<j Increase in Relative Risk for Vietnam Veterans in General Study Year 1 Control Group 'J: J' Pe °f Participant Soft Tissue Sarcoma Kasal & fta Liver Controls Number2 Prevalence of Vietnam Veterans 0.075 O.JOO 0.050 0.57 0.82 0.37 0.37 0.45 0.67 0.30 0.30 106 331 42 42 325 0.66 0.90 0.43 0.43 Study Yea- 2 Number2 Soft Tissue Ly;.;phoma I<'3fi.?l a J.^sv Li'-er Control.'3 Control Group Prevalence of Vietnam Veterans 0.050 0.075 0.100 0.70 0.92 0.47 0.47 212 662 85 85 650 0.83 0.98 0.58 0.58 0.90 0.99+ 0.66 0.66 Study Year 4 o Number Sort Ti&sue Sarcoma Liver Controls Control Group Prevalence of Vietnam Veterans 0.050 0.075 0.100 0.84 0.98 0.60 0.60 319 993 128 128 975 0.94 0.99+ 0.73 0.73 0.97 0.99+ 0.81 0.81 ( Study Year 4 Number Loft Tissue .Sarcoma Ly^phoma '•,a:-ial I I^ Liver Controls 425 1324 170 170 1300 o Control Group Prevalence of Vietnam Veterans 0.075 0,100 O.C50 0.92 0.99+ 0.70 0.70 0.98 0.99+ 0.82 0.82 0.99+ 0.99+ 0.89 08 .9 �28 Table VI (continued) b) 2-i:old and 5-fold Increases in Relative Risk Under Assumption of 7.5% Control Group Prevalence of Vietnam Service and 3 levels of Possible Agent Or a age Ex p osur e Among Vietnam Veterans (Study Year 4 only) 2-fold Increase iu Relative Risk For A^nt Orange Exposed Vietnam Vete_rans_ Type of Participant Soft Tissue Sarcoma Lymphoma Nasal & Nasopharynaeal Liver Controls Number 2 425 1324 170 170 1300 Possible Prevalence of Agent ,0_range_ Exposure Among Vietnam Veterans " 0.113 j).25 .°_'J12. 0.33 0.49 0.23 0.23 0.62 0.85 0.41 0.41 0.85 0.99 0.61 0.61 3-fold Increase in Relative Risk for Agent Orange Exposed Vietnam Vc^t^jran.-j Type_ ol: Participant Soft Tissue Sarcorm Lymplio'iid Nasal f, Masopharyngeal liver Controls N^E^r '£.2 425 1324 170 170 1300 Possible Prevalence of A;<ent: Or an 5^ e. }|£:p_osure Among Viatnam Vncarans * 0710 0.25 0.50 0.96 0.99+ 0.81 0.81 0.99+ 0.99+ 0.98 0.98 0.99+ 0.99+ 0.99+ 0.99+ Vower calculations with 1-tail, alpha - 0.05 by method of Casagrande. JT, Pike MC: An improved approximate formula for calculating sample si".es lor coiwaring two binomial d i s t r i b u t i o n s . liiometrics 1978;34:4S3-6. Estimated number of participant s �29 4. Pretests and Pilot Studies Ageiit Orange and Vietnam, Experienc.e Stiulies 'fwo major categories of procedures need to be assessed before the mala studies begin. First, there are a number of issues involving the manipulation of military records which need more work. Second, there is the matter of locating study subjects, securing their cooperation, and assessing the various study instruments (questionnaires, examination and laboratory protocols). The failure of any of the proposed procedures in preliminary tests will require revision of the procedures, and, if major failures are identified, outside consultation and peer review of new proposals. All proposed study procedures will be tested in a series of interrelated pilot studies and pretests. For the purpose of the discussion here, the term "pilot" study will be reserved to refer to the final process of assessing participation rates and evaluation of interview and examination Instruments • just before the start of the main cohort studies. The term "pretest" vi.1.1 bs used t.o refer to evaluations of all other procedures. It might be desirable to do .formal and complete pilot studies for each of the three proposed studies. However, because such an approach would unnecessarily lengthen the tir.:« required to complete the two cohort studies, CDC recommends that procedures be tested with a series of related "pretests" and "pilot" studies. In tlv>sr« situations where one among several alternative procedures clearly seems ro b« the method of choice, only that method will be pretested and the other ;il.U;rnatives tried only if the preferred choice fails. la other inst.mc:».'.K, there may be no clear preference and then more than one procedure will 1»f. pretested. Tho (V'.ueral approach for the pretests v?ill be early and close nonitoring of r:i rrinscribed aspects of the study procedures. Several pret?.scs of pror.e-Iures which would be sequentially applied in the main studies can be done sirriul t.-niHously. It is obvious that much time could be saved by using this approach. On the other hand, if problems are identified there would ba minimum delay and relatively little work necessary to repeat the process u-^ing corr*-r-t:cd procedures. Moreover, if no major problems are identified then the data >ynerated during the pretest could be used for the next pretest step or, for son* procedures, the processes judged to be successful in pretests could be ui;.*d straight aw>iy for the main studies. An example of the pretest approach is the evaluation which was done t.o assess Che locatability of male veterans, and the plans for making the sam-3 sort oT evaluation for female veterans. Tlie AAOTF transmitted to Ci)C identifying information for some 840 male veterans and CDC sent the information to the IRS to begin the locating process. The veterans used for this pretest were chosen because they were attached to twu units that tha AAOTI' had worked with previously (1st of the 9th and the 31st Engineers), The AAOTF had the names of the individuals who served in these units in 1967-1968 at hand, and only needed to request the personnel records from the St. Louis records center in order to obtain such items as SSNs and nan-es and addresses of relatives. IRS was able to provide locating information for 754 (89.8%) of the 840 veterans identified for CDC by the AAOTF, and of the 754 CDC was able to confirm locating by contact with directory assistance for 502 (66.6%) of the individuals. Thus, it appears that approximately 60% of subjects will be locatable through initial record check and the telephone system; contact of the remaining 40% will require additional system checks (e.g. SSA) including vital status determination. Clearly, "field follow-up" will be necessary to contact some of the subjects, and a subset, the size of which is currently unknown, will be unlocatable. The types of additional systems checks, erctent of field work necessary, and the probable siza of the unlocatable group will be determined during the pilot study. �30 The pilot study would be an integrated test of the contractor's locating systems, interview procedures, and the questionnaire itself. The pilot vrill involve approximately 550 veterans and require three months to complete. Results of the pilot would be reviewed continually, and changes in procedures and in the instrument could be incorporated as the need was discovered. Thus> at the end of the pilot period, data collection could begin immediately in the main study cohorts. Therefore, this request for data collection approval is for both the pilot and main studies. All changes in the questionnaires and collection procedures would be forwarded for review before main study data collection began. Military Records Pr e te s ts Because AAOTF has had extensive experience in working with records from the Vietnam era it is not expected that major problems will be discovered in the area of records manipulation. Even so, a more comprehensive test of the proposal to derive a sample of men for the Vietnam Experience study from the St. biuis records center was conducted to evaluate any problems which might arise in attempting to make the non-Vietnam veteran cohort match the Vietnam cohort in regard to calendar years of service. To this end a pretest sample of 241 Vietnam veterans and 322 non-Vietnam veterans was chosen. No serious problems were identified with the procedures. The scruples of veterans gathered during the pretest can be used as a part of the pilot study. Much work needs to be done with the records which will be used to classify exposure. While abstracting such data as daily unit locations :Lo apparently sinpl>?, at least for those familiar with the records, so little actual work in this nv.uni has been done for the purpose of assessing herbicide exposure it nsist !.••;• considered a relatively untried process. Rather than incorporate this ph-is* .into a formal pilot study, it is proposed that tha process be evaluated by cf">;>tant monitoring during the preliminary unit selection prce.'iss whsn th-.-. locHLumr; of tha 50 battalions are Identified. Even less experience; has been accrued in the process of checking troop locations against the herbicide records. In particular, the schemes proposed in this protocol for scoring herbiciiirt encounters have not been tried and their usefulness is unknown. Two priests of these schemes will be mad-'i. The first pretest will take place when r.he randomly selected units from III Corps are evaluated for the purpose of r-3tiking them on the herbicide encounter ccores; if cherrf appear to be no probV'ws at this stage, then CDC will have the AAOTF immediately proceed to the next step of the study, which will be the choice of individuals for Lhs ruin studies. Later the encounter scoring scheme will be tested again Cor individuals. Location Rate, Participation Rate and Instrument Aa mentioned above, some parts of the evaluation of the locstability o? t\r? cohort study subjects are now underway. This will continue as a. part of tha pilot study. Besides providing more information about locatability, the cohort pilot study will give information about expected main study participation rates and about possible difficulties with the interview instrument and examination protocol. The pilot study will be nearly a main study in miniature, the major exception being that the proposed selection process for the Agent Orange study cohorts will not be used to choose any of the pilot study subjects. As mentioned above, the subject selection process for the Vietnam Experience study provided 563 veterans eligible for the pilot study. Rather than wait for the process of ranking the companies in the 50 battalions from III Corps to be completed before selecting a pilot sample for �3.1 the Agent Orange study, CDC recommends another approach to save tine. It is proposed to sinulate the Agent Orange main study through tha use of 400 veterans who will be chosen from among the 110-120 combat battalions which were stationed in III Corps during 1967-1968. The selection of these pilot study veterans will involve tha initial random selection of 10 companies from the 110-120 battalions. From each of these companies, 40 randomly chosen men will be selected. Although the cohort pilot study will simulate the main studies, the results will be considered in two stages — an interview stage, which will almost certainly be completed first, and an examination stage. If the interview stage proves to be successful, CDC will proceed with the interviews for the full study sa?nple.3 even though the results of the examination stage may not be available. As noted elsewhere, CDC is concerned that it may be difficult to reach an acceptable level of participation in the examination phases of the studies. Tlia Ranch Hand study group's enviable success in this regard is attributed in large measure to their treatment of their study subjects as "VIFs." CDC will attempt to duplicate this treatment. Since there nay be monetary factors \vhich in!" lueru-.e participation in the examination phase, CDC will test the effort, of recompensing the subjects for lost time; offering recompense uny help 1.0 raise paticipation or it may decrease it if the offer offends a sense of all•cuisra. In addition, the effect of travel to distant locations for the exanii i i i t Lons ;aay enhance or deter participation. If it appears that mor?. than ona *:.•;.unliving center will need to be used in the main studies, a test of the effect. <>f distance to the center will be made in the pilot studies. ,c>:.'1i'.o.t:£d CancerG Study Ti> • .'ial.vct-.^d Cancers Case-Control Study procedures will bs field tested :in 2-3 Sl'.KK centers using fewer than 9 cases of lynphoira. Only lyr.'phoma cacitiJ •rfil.1 li.j iis^d because of the rarity of the other "selected" cancers and CDC carniu! risk "wasting" them on a pilot study. Only 2-3 3KER cenr..-;rs will bo used r;o minimize the time required — CDC feels that aoce are not required becaiinn of its previous success with the Cancer and Steroid Hormone study, The i-.iai.n purpose of a pilot study will be to evaluate the participation r*ite of in!.>•:•! aged 30-49 and the interview instrument. The work done by the AAOJ.'F on sf.ovlfig herbicide exposure likelihood for CDC's birth defects study is considi-.Tvii] :i valid surrogate for an assessment which could be done specifically for this study. 5. jS t-a t Is ti cal Design Th'^. statistical aspects of these studies were dealt vith by J. David Eru-.lc.'-ior), DJj.S., M.P.H., Ph.D.; Peter M. Layde, M ! . M.Sc.; and Matthew M, .), Zac'c, H.D. , M.P.>[. These individuals are all affiliated with CDC's Chronic Di^easv-s Division and may be reached at (L'".L'S) 236-4072. The identity of the data collection agencies is unknown at this tiu:-:-, since it is planned to contract for these services, and the competitive process is not complete. Data analysis will be performed by Center for Disease Control staff under the direction of J. David Erickson; Dr. Erickson ruy be reached at (FTS) 236-4068. Doc. 354ON � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 064 Folder The folder containing the original item. 1729 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Description An account of the resource <strong>Corporate Author: </strong>Centers for Disease Control Date A point or period of time associated with an event in the lifecycle of the resource 1983-11-01 Title A name given to the resource Supporting Statement: Epidemiologic Study of the Health of Vietnam Veterans Subject The topic of the resource Agent Orange Exposure Study Vietnam Experience Study Selected Cancers Study study protocol ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/ebf676b34455fb4108114ee372637d95.pdf db848ad79c1e4a031f02ffb0c9220580 PDF Text Text °1736 Item ID Number Author CorpOratB Author Agent Orange Projects, Chronic Disease Division, Cente RODOrt/ACtiClB TltlB Typescript: Project Description for: Epidemiologic Studies of the Health of Vietnam Veterans, July 29,1985 Journal/Book Title Year 000 ° Month/Day Color Number of Images n 2 Descrlpton Notes Monday, June 11, 2001 Page 1737 of 1793 �CENTERS FOR DISEASE CONTROL CENTER FOR ENVIRONMENTAL^HEALTH _- AGENT ORANGE PROJECTS -. .. - 1600 Clifton Road, N. E., Atlanta, Georgia 30333 Project Description for: EPIDENIOLOGIC STUDIES OF THE HEALTH OF VIETNAM VETERANS. (THE "AGENT ORANGE PROJECTS.") The overall investigation includes three separate but related components: 1) Agent Orange Study. (Study of the long-term health effects of exposure to herbicides in Vietnam.) 2) Vietnam Experience Study. (Study of the long-term health effects of military service in Vietnam.) 3) Selected Cancers Study. (Study to determine the risks of specific cancers among Vietnam veterans.) BACKGROUND BetweenRugust 1965 and February 1971 approximately 11.3 million gallons of the herbicide "Agent Orange" (so named because of the orange markings on the drums in which it was shipped) were sprayed over much of South Vietnam in military operations designed to deprive the enemy of cover and food. A chemical contaminant, 2,3,7,8-tetrachlorodibenzo-p-dioxin, more often called TCDD, or simply dioxin, was created during manufacture of and contained in the Agent Orange which was sprayed. Dioxin has been shown to be a highly toxic substance. In January 1978 the Veterans' Administration (VA) received the first of what was to become many claims from veterans who felt that their current health problems had resulted from their being exposed to Agent Orange while serving in Vietnam. In January 1979 the U.S. Congress enacted legislation (Public Law 96-151) directing the VA to design and conduct an epidemiologic study to determine if exposure to Agent Orange had caused long-term adverse health effects in Vietnam veterans. In November 1981 the scope of the study was expanded (by Public Law 97-72) to include other factors in the "Vietnam experience," including medications and environmental hazards or conditions. In January 1983 the responsibility for designing and conducting the investigation was transferred from the VA to the Centers for Disease Control (CDC). In May 1983 CDC scientists completed detailed guidelines (protocols) for the Agent Orange and Vietnam Experience studies, recommending that a third investigation be conducted at the same time to determine the risk of Vietnam veterans developing selected types of cancers. Public "Notice of Research Project Initiation" was published in the Federal Register on March 13, 1984. DESCRIPTION; AGENT ORANGE AND VIETNOT TEXPERIENCE STUDIES11 Although both of Wese""hT.istoHcai(" 'of~nreEr5sp^cHv7e7 studies are in some respects similar, each has a separate purpose. The Agent Orange study is designed to find out if troops who were exposed to the herbicide during service in Vietnam have suffered long-term adverse health effects as a result of that exposure. The Vietnam Experience study is designed to demonstrate whether or not thcro is any difference in the health of veterans of the Vietnam era who nerved in Vietnam compared to the health of veterans who served in other countries during the same period of time. The studies require the cooperation of a large number of Vietnam era veterans willing to be interviewed about their health status and experiences before, during, and after those years. TO ENSURE STATISTICAL ACCURACY, NO VOLUNTEERS CAN BE ACCEPTED AS PARTICIPANTS IN THE STUDIES. Participants are selected following scientific guidelines established by the research protocols. With the help of the Department of Defense and other agencies, CDC will identify a .minimum-flf 30,000 qualified veterans to participate in the studies: 6,000 in each of five separately defined groups~T3t "cohorts." The five cohorts are to be made up of veterans who: 1) Served during 1967-68 in a specified area of Vietnam, and were likely to have been exposed to Agent Orange. 2) Served during 1967-68 in the same area of Vietnam as cohort 1, and were less likely to have been exposed to Agent Orange. 3) Served during 1967-68 in another area of Vietnam than cohorts 1 and 2, and were not likely to have been exposed to Agent Orange. 4) Served in Vietnam during 1966-72. areas. 5) Served during 1966-72 in countries other than Vietnam. Randomly selected from all Data for the Agent Orange investigation will be gathered from cohorts 1, 2, and 3. Cohorts 4 and 5 will provide data for the Vietnam Experience study. AGREEING OR DECLINING TO PARTICIPATE IN THE STUDY WILL HAVE NO EFFECT UPON BENEFITS A VETERAN WAY BE RECEIVING OR TO WHICH HE WAY BE ENTITLED IN THE FUTURE. All information given by each veteran will be held in complete confidence. The names of the participants will never be associated with their answers in the statistical summaries studied by scientists. Names and other identifying information, such as addresses and social security numbers or service numbers, will be kept in a separate file that no one will have access to but the U.S. Public Health Service and the private research firms working on this study. No other researchers or government agencies, including the Veterans Administration and the Department of Defense, will be able to learn if a veteran participated or what his answers were. This promise of confidentiality is guaranteed by Federal laws—42 U.S. Code 242(b), (k), and (m). Unless the veterans gives written permission to CDC to release personal information, no one, including the veteran s family, will ever be able to get the personal information provided by the veteran. The interview takes about 40 minutes and is conducted by telephone by CDC's contractor, Research Triangle Institute (RTI), Inc. Veterans who are selected to be called by RTI receive a letter from CDC telling them to expect the call. From those being interviewed, approximately 2000 veterans from each cohort will have been preselected for the medical examination component^ the study. The RTI interviewers have no control over which veterans will be asked to take the medical exams. �»,0nl,y ^veterans who have already been interviewed by medical exams which will take 3 days to complete. veteran selected will receive a letter explaining Lovelace Medical Center asking when he can come convenient. io_. themselves. RTI will be selected to be asked to take the Several weeks after being interviewed, each the examinations and a telephone call from to Albuquerque. Veterans can select dates The 10,000 medical examinations are being conducted at non-hospital clinical facilities specially constructed for this project by another CDC contractor, the Lovelace Medical Foundation, in Albuquerque, New Mexico. All examinations are being done at the same place to ensure that standard testing procedures are used. The examination includes about 60 physical, psychological, and laboratory tests. Blood and urine samples are required, but no tests are included that most persons would find painful. Participants can refuse to take any test or to answer any question. Veterans who complete all the tests receive a $300 stipend. Veterans' expenses for_trayel..to, and from Albuquerque, food and lodging, etc., will be paid by the government. Veterans will "stay in private rooms at a first-class downtown hotel and have their evenings free. Each room will accommodate up to four persons without cost to the veteran. (The government cannot pay for family members' travel or food.) Physicians and other health providers working on the CDC studies will not provide any treatment for individuals. If a veteran's medical examination indicates the possible existence of a problem of any sort, the veteran will be advised immediately and encouraged to ,seek treatment from the VA, private, or other sources of medical services. Veteran interviews for the CDC study began in September 1984, and will continue until about October 1987. The first medical examinations were conducted in March 1985. All examinations are expected to be completed by about January 1988. RTI, Lovelace, and other non-government research firms have been contracted to collect the data for thesestudies. These firms are monitored closely by CDC officials. All analysis and interpretation of data is done by CDC. .&fir>~ "\~3e-/ -* DESCRIPTION; SELECTED CANCERS STUDY There is some scientific evidence that exposure to herbicides may increase the risk of several serious, but relatively rare, cancers in workers in industries which manufacture or use similar products. Because these cancers are so infrequently seen, the 30,000 veterans in the other study cohorts do not offer a large enough sample population upon which to base this investigation. Instead, two other groups will be studied in a "case-control" investigation. Because of the design of this study, veterans and non-veterans will be included in both the case and control groups. The tumors selected for the study are: lymphoma, soft-tissue sarcoma, nasal and nasopharangeal cancer, and primary liver cancer. Other types of tumors may be added to the study later. The first (case) group in the Selected Cancers Study will be made up of male patients who have actually had these tumors, and who could have been in the military during the Vietnam conflict. * The^jjpecpnd ^(cgntJEOlJ)-j group uiJ,J, inpJLude roen.ji&.Mia 9fme. «QO and .fi-pOL tho .eanw current geographic "area as" the case cohort, but ""without the ibumbrs. Using information from interviews and military records, CDC will determine which men in both groups are veterans, which veterans served during the Vietnam era, and which veterans may have been exposed to Agent Orange. Comparison of data collected from both groups may indicate significant differences in their risk of these cancers which could be associated with military service, service in Vietnam, and exposure to Agent Orange. INVESTIGATION RESULTS The exact rate of progress of epidemiological studies of this size cannot be forecast. Collection and analysis of ths large amounts of data needed for scientifically valid findings takes time; particularly when so many thousands of veterans must be identified, located, interviewed, and examined. CDC will report on each component of the study when it has been completed. Final reports on the Agent Orange and Vietnam Experience components are expected by September 30, 1988. The final report on the Selected Cancers Study component is expected by September 30, 1989. CDC hopes that these studies will provide answers to many of the important questions being asked about Agent Orange and other factors related to service in Vietnam. But, as in every epidemiologic investigation—no matter how carefully designed and professionally conducted—the possibility exists that definitive answers to some questions may never be found. 07/29/85 � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 064 Folder The folder containing the original item. 1736 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Description An account of the resource <strong>Corporate Author: </strong>Agent Orange Projects, Chronic Disease Division, Center for Environmental Health, CDC Title A name given to the resource Typescript: Project Description for: Epidemiologic Studies of the Health of Vietnam Veterans, July 29, 1985 Subject The topic of the resource Vietnam Experience Study Agent Orange Exposure Study Selected Cancers Study ao_seriesIII