1 15 29 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. 069 Folder The folder containing the original item. 1864 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/. Creator An entity primarily responsible for making the resource Greenwald, Peter Vito Logrillo Description An account of the resource <strong>Corporate Author: </strong>State of New York Department of Health, Office of Public Health Date A point or period of time associated with an event in the lifecycle of the resource 1980-11-01 Title A name given to the resource Proportional Mortality of Vietnam Veterans, Other Veterans and Matched Controls in New York State, Exclusive of New York City Subject The topic of the resource state-funded Vietnam veterans study mortality trends cohort study ao_seriesIII 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. 063 Folder The folder containing the original item. 1718 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/. Creator An entity primarily responsible for making the resource Erickson, J. David Title A name given to the resource Letter: from J. David Erickson to Recipients of Draft Protocols for Epidemiologic Studies of Vietnam Veterans, June 3, 1983 Subject The topic of the resource Agent Orange Exposure Study study protocol cohort study ao_seriesIII 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. 063 Folder The folder containing the original item. 1717 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/. Creator An entity primarily responsible for making the resource Taylor, Charles S. Title A name given to the resource Typescript: Science Today: Major Agent Orange Study Planned, May 17, 1983 Subject The topic of the resource Agent Orange Exposure Study Agent Orange controversy cohort study CDC Birth Defects Study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/a7731d0a2d5f7c9341f44f1ed61c9dd7.pdf b7dc609179a75b372b614017d158816d PDF Text Text Hem 10 Number °1716 Author Jemison, Terry Corporate Author Report/Article TltlO Agent Orange: Exposure Tackled' JOUrnal/BOOk Title U.S. Medicine Year 1983 MODth/Day January 1 Color Number of linages n 2 Doscrlpton Notes Monday, June 11, 2001 Page 1717 of 1793 �-.«,.. "\ Gonorrhea Vaccine By Nancy Tomlch *1 ,,, •i n Independent 'Notional Newspaper for physicians e WASHINGTON—A gonorrhea vaccine 'developed jointly by Waller Reed Army Institute of Research and the University of Pittsburgh will undergo its first clinical trial early this year. • ; The vaccine, which has'been "in the bottle" for a year and a half, will be tested among 5,000 Army, Navy, Air Force and Marine volunteers stationed in Korea in a double-blind,'randomized placebo study. . .1 .£, Half the group of volunteer subjects Congress Redirects Funds Expansion Of CO MISS Halted -'• 0 •..' ."'.'"' .'•/,:' •.••;*<'', J. ' • . • . will receive the vaccine in two doses of 100 micrograms each,- given two weeks apart. The other half will be injected with a placebo. The study subjects will be monitored for eight weeks, according to a Defense Department source familiar with the protocol, after which each will be cultured. / When the study code is broken, the source related, it hopefully will be possible to tell whether the vaccine conferred protection against gonorrhea, which is a local infection of the genital area. This will be accomplished simply by determining who has an infection and who does not, and whether these individuals received vaccines or not. '"We won't know if anyone has been exposed, to the disease," the source related/In fact, we will be discouraging these people from going out and having sex. No one would recommend they do that. "We firmly expect the rates of sexual The language placed in the CR by the <- "UTON—As it wound up a maraHouse Appropriations Committee lifts ^ 'isigned ?o keep 'govern- . a S35 million limitation on funds for ~s'- 2"; operating, Congress last .; medical data processing that earlier had . "• ,Vd language prohibitingv been placed in the fiscal 1983 appropriatne 0.1101,,:*.. jf the Veterans Administions bill for VA. ; tration's COMISS f pharmacy-system. : ,.; Specifically, the continuing resoluIn a continuing resolution passed at {Continued on page 26) tion directs: the 25th hour—loo'late to prevent agen•Stopping all work on the centralized cies from shutting down for a few hours < COMISS system "except the operation on December 20-^CongresS extended funding for all agencies for which reguand maintenance of the currently operating outpatient pharmacy system at lar appropriations bills had .nor been existing locations.* passed. V,- . <'*• •> ••}: .-'.•?'' ''-!:r'3'•Transferring 69 of the positions (52 Although an appropriations measure ;! FTEEs) assigned to the Office of Data had been passed for the VA, language '' Management and Telecommunications halting COMISS was placed in the con- :•; for development of COMISS to the linuing resolution, ;or "CR" as it,was;>;.. Department of Medicine and Surgery. referred to in thosefinaldaysofthe97th, J, Congress, by the House Appropriations ' "This will leave 11 positions, including not to exceed one in the central Committee,' whose members directed that d:centralized field-developed -' be given the entire increase on top of the office, to operate and maintain the outmodules be expanded instead. ,..; ;' special pay they now receive, have their patient pharmacy program in the ChiThis language was accepted.by both ' special pay* adjusted in proportion to cago area." •Limiting the Office of Data Managethe full House and Senate, meaning the the full amount of increase in base pay COMISS system now Is limited,to the or be given an amount of increase that ment and Telecommunications to 1,934 Electronpholomlcrograph of Neisseria , ', Chicago-area hospitals in which it falls somewhere in the middle of these FTEE In fiscal 1983. (Continued on pair }») gonorrhea. two already.is operating. ;_;; 'i-s^g^.4..',^,. - extremes. In adopting' the.'CR'i whi'ch^ixieho's ufe. ,. •.• • • . -, until September 30;;i983, Congfessalso : *"':J H*' ilHOMry*** K Ytl/>CIIPA " lifted the pay cap for federaj^xecutiyesJ« n , \J\ClA\tL\Z* i En AUUO U 1C ; Pay for members of the SeniorjExecu*"^ * '' ' ' 4 •% tive Service now will rise to a celling of $67,200 from the current one of $58,500., CDC Now Faces s Dosage For employes in the GS 15 step? level;., Challenge Of througn GS 18, who have been capped ' Calculated By at $57,500, the ceiling now rises to Cohort Selection Dob Formula $63.800, an increase of 11 per cent.. In VA, there was question whether WASHINGTON—A ticklish task of defining exposure to agent orange soon will physicians would receive the entire II WASHINGTON—Finishing touches ' are officially rest on the shoulders of scienper cent increase or none of: it.,'/, , being integrated into a Defense Departtists at the Department of Health and An "option paper" was being circu- ment-plan for gauging ground troops' Human Services. lated within the agency to obtain view- exposure to agent orange, a formula It is a question that was not so critical points as to whether physicians should that has been under refinement for a for the first federal epidemiological The DoD formula has yielded a quan- study already begun, that involving the :; Bart Hull,' co-chairman of a high-level .,-iinieragency work group 'on agent titative dose estimate for the soldier in Air Force Ranch Hand Unit, a clearly : • •orange appointed by President Reagan, thick jungle, directly under a Ranch defined population of flight crews that said the group has set up a science sub- Hand run: a maximum dose of .0010752 sprayed the herbicide over Vietnam jungle and cropland. ' committee specifically to coordinate the ounces of agent orange per soldier. The most controversial part of the Now, attention is turning to a new . .refinement of the government's leading ;;:, , - 'f^'plan; or "concept paper," for calculating DoD concept paper for the major and much larger study of the ground ground troops study is a three-tiered troops who were underneath the jungle WASHINGTON—Faced with a tedipus^exposure. balance between physician morale and ',''/• The Agent Orange Working .Group system proposed to choose cohorts of canopy. After HHS officials said at a hearing relentless waves of White House fiscal 'was scheduled to meet in mid-December, high (or low) exposure based on bat' in September they could have proceeded conservatism, the Veterans Administra- , .'and a Defense Department source said talion location. "Because of the concerns expressed with a health study of Vietnam ground tion has taken a middle road—giving" the'methodology for estimating agent doctors more money, but not as much as -.orange-exposure was "just about ham- by the Veterans Administration and troops faster than the Veterans Admin! (Continued on pate 7) istration has acted to design one, key it might have. , '. ,*" j*-'ci?4' -' :mered out." members of the committee moved to let The new formula officially was sighed them prove it. into policy two weeks before Christmas, I A letter from the House Veterans offering relief to thousands; of physiAffairs Committee convinced VA to cians in the upper pay grades .who. relinquish management and supervision because of a variety of federal policies, ; Corps Changing of the retrospective cohort study. earlier had been denied the full'4'per When an interagency agreement is cent pay raise for federal workers the - The legislation authoring the signed this month, it will be up to an President announced in October.''"; NATIONAL National Health Service Corps HHS agency, the Centers for Disease The VA decision, which was made expires al the end of 1983, Program HEALTHJ Control, to resolve the question of retroactive to October, when the current; L director Billy Sandlin predicts the exposure. SBRVICB fiscal year began, could cost the agency corps will replace its scholarships CDC must decide if the current plan $21 million, or S13 million more than with loans or other mechanisms. CORPS for study subject selection according to the minimum needed to meet the new Report, page 2. exposure, recently criticized by the legal requirements of the President's 4 National Academy of Sciences, will be per cent raise for federal workers. modified at all—or retained at all. The cost became particularly impor- t VA Follows Mid-Course On MD Pay • •< '(Continued on page 29) f Continued on poge 8) �, .'-JANUARY I, 1983 ' 1 U.S. MEDICINE Formula To |Ca/cu/afe' Agent 0/^iigep)psages Other variables and influences in soldiers' exposure to agent orange in Vietnam have been identified by Dr. Bricker i ; •There is no way to estimate dosage from the inhalation of the airborne droplets of herbicide, each smaller than 300 micrograms. ' •Agent orange was "utilized in considerable quantities around bases and along lines of communication." Less rigidly controlled than Ranch Hand UC 123 missions, helicopters often sprayed the herbicide at low altitudes in areas cleared of high trees, increasing exposure concentration due to rotor blade downwash, lack of tree foliage absorption and close proximity to troops. Such helicopter spraying not only included camp perimeters but armored personnel carriers and guard towers. •Other spraying was done from trucks and with finely atomizing backpack sprayers. Fire base perimeters were sprayed about every five weeks with the spray frequently drifting back over population centers with "a much closer and far more concentrated exposure to herbicides than for troops under a dense jungle canopy being sprayed by CI23 aircraft," Dr. Bricker said in his report. •At low altitudes, the Ranch Hand flights occasionally would jettison the herbicide under low emergency conditions, which could create a ground exposure rate 500 feet below of as much as .1424 gallons per foot. The emergency jettisons, which were announced by Health and Human Services secretary Richard S. Schwcikcr in September, 1981, when he learned of them, would make a good component of the criteria for selecting ground troops for the VA epidemiological study. Dr. Bricker said. However, the first criterion for selection of the exposed cohort would be close proximity to Ranch Hand sprayings. Ten battalions in the spraying areas on multiple occasions could yield 24,000 personnel, and the VA had anticipated needing a "highly exposed" cohort of only 12,000. CDC plans for the size of the study cohorts have not yet been determined. In what Dr. Bricker calls "step 2 criteria," the records of the base camps and fire bases occupied by each of the 10 battalions would be reviewed in detail to determine when camp perimeters were sprayed with agent orange during a oneyear period. Those battalion-size units meeting both step I and step 2 criteria would be examined for a third criterion, that units of the battalion were encamped or operating within 2 km. of one of the low altitude jettisons. Similarly, Dr. Bricker proposed using the criteria to test for the "proposed unexposed Vietnam combat cohort." Ten battalions whose units operated exclusively in non-sprayed provinces would be selected, again in hopes of a total troop complement of 24.000 soldiers. After selecting units not meeting "criterion I," or proximity to Ranch Hand Unit spraying, the second criterion also would be applied in reverse—finding units not exposed to any local perimeter herbicide spraying. The third criterion in reverse, not being located near where loads were jettisoned in emergencies, is easier to meet, he said, since those provinces were not in the heavy combat spray area. !&,;.,.^ »•<//«>« wt it] • ;•• .:^'i};''i.;|;;|^erans'§dministration officials, ...-.'•.^veterans.-, :: ! 1 ''faced ••"•'•-~'--i!-1'"•""•""linn:/1""k many members ohhe (working group's) ' ''-"- with potential compensation costs science panel as to the precise methodol- of billions of dollars, have been cautious ogy for the selection of cohorts', a mpra-' iii such extrapolations and insist that torium on (he actual selection process for only positive data from studies specifithe pilot study (to test methodology) cally of humans can validate that was requested pending refinement of a theory. / • ' -,. The Bricker exposure estimate of.166 valid methodology," Kull said. •''''. . Kull, an intergovernmental affairs gallons per acre for Vietnam groundspecialist at the agency that has lead level concentrations is an estimate responsibility for managing the working assuming several factors that were not group, the Department of Health and always present, such as dense, triple canHuman Services, said the working opy jungle. If an area had been sprayed earlier, group's concerns were similar to those expressed by the National Academy of and the exposure in question were to involve a re-spraying mission, the Sciences (see related story). The review, he said, will not impede absence of foliage (due to the first treatselection of cohorts, but the refined ment) would increase the proportion methodology will assure "valid, reliable, penetrating down to the 8-foot zone. timely and relevant data that veterans According to the Bricker report, can trust will produce reliable answers troops under "sparse canopy or relaon this vital issue." tively open forests" might have gotten a One question looming over applica- dose of agent orange of up to .0181 ountion of the exposure estimation plan in ces per soldier. In the dense jungle setan epidemiological study is the potential ting, that dose was estimated at a confounding influence of the general maximum of only .0010752 ounces per United Slates population's exposure to soldier. agent orange components. In" a memorandum to the science Scientists are concerned about veter- panel explaining his calculations. Dr. ans' earlier domestic exposure to dioxin, Bricker assumed that the agent orange the toxic chemical federal scientists have probably had a 2 parts per million condetermined to be carcinogenic, tcratc- centration of dioxin (TCDD). With only about .0005376 ounces of genic and fetotoxic in some animals and which has been implicated as the likely agent orange per square foot reaching villain in any long-term agent orange the 8-fobt'ground zone, Dr. Bricker health effects. said, "five ten-thousandths of an ounce Agent orange was a 50-50 mix of two per square foot is a very small amount if herbicides, the chemicals 2,4-D and it contained 2 pjvn of TCDD." 2,4.5-T. It is the latter whose manufacturing process inadvertantly introduces dioxin as an impurity into the herbicide. Over 30 years, the herbicide 2,4.5-T grew into wide use in American agriculture, the forest industry, and even as a popular lawn chemical used by home gardeners, such as in "weed-and-feed" 1.000 Gallons of Agent Orange Sprayed fertilizer mixtures sold at the corner hardware store. The architect of the much-touted agent orange exposure estimate methodology, Jerome G. Bricker, PhD. speculated in a report on his model for the major gVound troops study, "It may be impossible to find any group of persons who have not had some exposure to dioxin if they are older than 10 years." Dr. Bricker has estimated the typical agent orange exposure rate at the "ground zone," from the ground up eight feel into the jungle, as 0.166 gallons per acre. However, in the U.S., the customary agricultural use—before a 1970 banwas a rate of two gallons per acre, 870 Gallons Reach Top of 12.3 Million Sq. Ft. Jungle Canopy according to Dr. Bricker's report, which Concentration: .0000707 Gallons/So.. Ft has become the cornerstone for estimating exposure accepted by the Cabinetlevel work group on agent orange. Dr. Bricker said there actually are three methodologies under development. Kull said there is a "misunderstanding in some quarters" that they are designed to be interchangeable, but they actually are separate and are designed to meet different needs for different studies. Dr. Bricker, health legislation specialist at the Defense Department, is a DoD emissary to Kull's interagency working group. The powerful defoliant 2,4.5-T largely was phased out of American agriculture between 1970 to 1979 by the Environmental Protection Agency, which has canceled registrations for most such uses. In 1980 regulatory actions to formally cancel commercial 2,4,5-T permits, the EPA determined there is a reasonable basis to conclude that humans are likely to experience effects comparable to An interagtncy working group on agent orange Is refining a "concept paper" calculating ground troops'exposure that Includes TCDD's carcinogenicity in mice and this Defense Department exposure estimate. The concept paper, authored by Jerome C. Bricker, PhD, assumeslhls dost at the 'ground zone' from a C123 Ranch Hand aircraft flying at 150 knots and 150 feet altitude. rats. � 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. 063 Folder The folder containing the original item. 1716 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/. Creator An entity primarily responsible for making the resource Jemison, Terry L. Source A related resource from which the described resource is derived U.S. Medicine Date A point or period of time associated with an event in the lifecycle of the resource January 1 1983 Title A name given to the resource Agent Orange: Exposure 'Tackled' Subject The topic of the resource Agent Orange Exposure Study study protocol ground troops exposure cohort study ao_seriesIII 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. 063 Folder The folder containing the original item. 1715 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/. Title A name given to the resource Typescript: Point Paper: Herbicide Orange Exposure Concepts for the VA-CDC Epidemiologic Study of Ground Personnel: Unofficial Letters by Drs. Albanese & Lathrop to CDC Pis and Resulting Misconceptions, [nd] Subject The topic of the resource Agent Orange Exposure Study study protocol cohort study study criticism ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/31b0bbdccd229077597cd042b455ae57.pdf 09ea438ddace4088d7d1231096cb18a9 PDF Text Text °1714 Item ID Number Author Corporate Author KOpOrt/ArtlOlO TltlO Typescript: Cohort Formation for Epidemiological Study of Health Effects of Agent Orange, Fifth Draft, December 15, 1982 Journal/Book Title oooo Month/Day Color Number of Images Descripton Notes Monday, June 11, 2001 Page 1715 of 1793 �December 15, 1982 FIFTH DRAFT COHORT FORMATION FOR EPIDEMIOLOGICAL STUDY OF HEALTH EFFECTS OF AGENT ORANGE 1. Reasons for Cohort Formation To determine whether exposure to Agent Orange, a phenoxy herbicide used in Vietnam, adversely affected the health of American military personnel, the physical and mental health of a group or cohort of exposed veterans must be compared with the health of a cohort of relatively unexposed veterans. If such groups can be accurately identified, a cohort or follow-up (historical prospective) study design is both possible and desirable. For greatest validity, the two cohorts should resemble one another as closely as possible except for their members1 contact with Agent Orange. 2. Problems of Cohort Formation Estimating exposure is the crucial step in forming the exposed and nonexposed cohorts that are the keystone of this study design, tfo records were kept of an individual's exposure to Agent Orange, however, because it was believed to be a safe herbicide. Records do exist, for most of the dates and places where the herbicide was heavily used and separate records exist for the daily location of many specific military units. It is possible also to indicate the dates each man was with his unit by reconstructing rosters of the men in units with adequate records. From these data it might be possible to estimate the extent of exposure for many individuals. The likelihood and intensity of exposure could be evaluated using at least the type and time of exposure, and distance from the point of herbicide application. The "type of exposure" includes such experiences as entry into an area recently sprayed by an aircraft, presence in a base camp during perimeter spraying, or being nearby when a plane dumped or jettisoned its. load of herbicide in an emergency. The "time of exposure" refers to the period between the herbicide application and the unit's presence in the area. �The "distance" refers to the proximity of the unit to the point of application. When enough information is available to determine the relative amount of exposure under various circumstances, the exposure at each episode can be "weighted" and the weighted values for each individual can be summed to give an "index of exposure". In an initial retrieval test of cohort selection, such an index can be applied to determine whether a heavily exposed and a non-exposed or minimally exposed cohort can be defined. If separation is obtained by criteria acceptable, to the scientific community as well as knowledgeable veterans' representatives, the essential cohort selection will have been shown to be feasible. 3. Development of Exposure Cohorts Of prime importance to a proposed epidemiology study of the effect of exposure to Agent Orange on the health status of Vietnam veterans will be a clear, thorough statement of the process used to identify study subjects. The procedure currently recommended by the Army Agent Orange Task Force (AADTF) has been designed to make this selection feasible, efficient and as unbiased as possible. To minimize the very large number of records to be reviewed and data to be entered a multistage selection process is used. The basic selection or "sampling" frame will be the military unit which will be characterized as "exposed" or "not exposed". Individuals will then be identified from selected units and further studied to assure their experience is correctly represented by the unit's classification. The retrieval test phase of this study will determine whether adequate cohorts of exposed and non-exposed can be defined. To accomplish this an audit trail will be maintained of all units identified, the proportion selected and the basis for exclusion of the non-selected units. Thus the retrieval test will also perfect the selection process to be used during the full study. 2. �For a meaningful comparison of outcomes among subjects "exposed" and "unexposed" to Agent Orange, cohorts ,should be as similar as possible in other aspects of the Vietnam experience. Thus, units should be stratified to reduce heterogenicity. It should be sufficient to select exposed and unexposed units from the same branch of service, in a similar type of unit, and operating in the same Corps. If possible, similar terrain would be a useful fourth criteria. In the first step in finding unexposed units, candidates are selected from among those that were operating beyond the reasonable maximum drift for Ranchhand spraying. It will be necessary to evaluate other potential exposures (e.g. perimeter spraying, dumps, etc.) for a final classification into exposed or unexposed units. The availability, completeness and accuracy of the various record systems is unknown at present and this will be studied during the retrieval test. However, the procedural guidelines or selection algorithm used by the AAOTF is to a large part based on assumptions about these records. This increases the likelihood the AAOTF staff will encounter unexpected variations. If this requires AAOTF staff to make eligibility decisions on units or individuals not anticipated in the algorithm, a brief anecdotal record of the nature of the problem and the convention adapted to solve it is most important. This will facilitate an evaluation of the cohort selection process for potential selection bias during the retrieval test and will determine the feasibility of the proposed epidemiological study. The following procedures are for selection of Vietnam service subjects with high and low likelihood of exposure to Agent Orange or other herbicides while in Vietnam and do not include the selection of a non-Vietnam cohort. The outlined steps have been included in the guidelines supplied by the AAOTF. The documentation requirements provide information to evaluate the feasibility of identifying adequate cohorts for the full-scale study. 3. �Specific attention is directed to the concept of "tracking time". An individual need only be in an exposed unit for enough "hits" to qualify as heavily exposed. Therefore the tracking time for these units would be just enough months to generate an adequate number of personnel exposed. They may be followed for a shorter period. On the other hand, the non-exposed or minimally exposed individual must be followed for his entire tour. An individual typically served exactly 1 year and could enter the unit throughout the year. Therefore, a non-exposed unit's tracking time must be more than 1 year to establish this category. How much longer again is determined by the number of non-exposed needed. I. Prepare maps of Vietnam with Ranchhand spray tracks (both defoliation and crop) for the period 1967-1968 (this has already been provided). II. Define geographically homogeneous areas with known ground troop activity with and without Ranchhand tracks during the 2-year period. This is the first step and defines areas which might be selected. The rationale for choosing selected areas must include areas with and without Ranchhand tracks, similar location (and if possible terrain), and the presence of ground troops. If possible, all four Corps of RVN should be represented. III. a. Obtain appropriate station lists for each area selected. b. Determine potentially eligible units which were operating in selected areas during the 2-year period. , All units identified from station lists are potentially eligible. IV. Develop "enhanced" HERBS Tape by including information from herbicide applications not identified from Ranchhand missions. This will include perimeter sprayings, road clearings, helicopter missions, aborts and other identifiable applications. It should not be . necessary to do nore than those areas previously identified in Step II above, but, it will be necessary to include an adequate tracking time based on the unit's exposure status in those areas to be studied. 4. �V. Determine availability and completeness of all eligible units' records of movement. A necessary part of evaluating the reliability of exposure estimates will be the assessment of estimated completeness of the available records for herbicide applications. Documentation of completeness will assist in this effort, especially the number and type of obvious errors and suspected inaccuracies. VI. a. Select units to be recorded as to daily movements. This is a crucial step, and documentation of how and why a unit (or units) were selected should include the eligibility of all units which were potentially eligible and might have been selected. The potential for the selection of adequate numbers of units (or subjects) for a full scale study may depend on records' completeness and availability at this step. Obviously only a small subsample (one or two areas perhaps) can be done in the retrieval test. b. Record to machine readable format the daily location coordinates for selected units for the entire tracking time. VII. a. Match daily location coordinates of selected units via computer with the "enhanced" HERBS Tapes to determine the number and type of "hits" sustained by each eligible unit between January 1, 1967 and December 31, 1968. Each "hit" will be defined and recorded according to all of the following four parameters for each unit: i) Date ii) Type of application (Ranchhand, Helicopter, Aborts, Ground, incidents) ( 3 ^ erf ^ ^ f iii) Time in days since application « 1 day) ( iv) Distance from application site in km ££=ir=km, 2 5. �b. List the "hits" for each unit matched and the weighted score. An Appendix illustrating an index or weighting scheme is in preparation. While it may not be possible to quantify exposure with certainty, at least a weighting rule may be developed to classify high and low exposure units. The last category in both time and distance parameters is intended to provide a buffer between exposure and non-exposure. A "hit" in this time-distance zone will not constitute an additional exposure in otherwise exposed units but will be considered a possible exposure in units to be classed as low-likelihood-ofexposure. Thus, "low-likelihood-of-exposure" units will not be considered to . have been exposed unless they have been within 1 km within 15 days or within 5 km within 1 day of a Ranchhand application target (if the retrieval test shows this is feasible). EorJjmts^gj^perted -t^FBe^xpo^etl'r^the time-and distance- --iv abov^f^to'^efAne ^a- .hik jaeed, jonlyJae-measured at 3 VIII. Select high and low likelihood of exposure units. This step will depend on selection via the weighting system developed for the previous step as well as the results from the retrieval test. The rationale for selection will be indicated and should include other criteria for the type of units chosen as well as herbicide exposure. Depending on the results of Step VII, it may be possible to identify "low likelihood of exposure" units with no "hits" of any kind. If this is the case, then a specific weighting system to discriminate between high and low exposed units will not be crucial for selection. "High likelihood of exposure" units can then be selected from among those units with at least several "hits" of various kinds (exclusive of. buffer zone hits). IX. a. Obtain daily morning reports for an adequate tracking time during the period January 1, 1967 to December 31, 1968, for each unit , selected in the previous step. 6. �b. Track individuals through service with their respective units during this time period and record presence on "hit" days for their unit. Some criteria for eligibility will have to be developed and documented, including the minimum number and quality of individual's "hit" days, reliability of records for identification of individuals during this step, and their minimum length of assignment to the unit. X. a. Obtain service records for selected individuals and verify service information obtained via morning reports, b. Record personal information from service record and include location of medical record for each selected individual. At this step, it may be desirable to retain the individual's identification as either exposed or unexposed during the period covered in these procedures. This should enable closer matching of the two groups based on individual characteristics obtained from the service record. It will be necessary to identify multiple tours of duty in Vietnam and perhaps other criteria pertaining to military service which will influence an individual's final selection into either of the Vietnam service cohorts. Documentation of these inclusions and exclusions will be necessary in order to assess bias in the final selection. Although most of these procedures are included in the guidelines supplied by the AAOTF, they are detailed here to identify the steps and type of documentation which will be useful in the evaluation of the exposure index and determination of the feasibility of selecting exposed and unexposed cohorts for the full scale study. Systematic departures from the procedures outlined here which may be necessary for the inclusion of subjects from each of the four branches of military service should be included in the documentation. 7. � 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. 063 Folder The folder containing the original item. 1714 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/. Title A name given to the resource Typescript: Cohort Formation for Epidemiological Study of Health Effects of Agent Orange, Fifth Draft, December 15, 1982 Subject The topic of the resource Agent Orange Exposure Study study protocol cohort study ao_seriesIII 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. 063 Folder The folder containing the original item. 1713 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/. Creator An entity primarily responsible for making the resource Albanese, Dick Title A name given to the resource Letter: from Dick Albanese to Dr. Welty, November 12, 1982 Subject The topic of the resource Agent Orange Exposure Study study protocol cohort study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/bbb9b7c4792cc1108c05bd9b3d325692.pdf 0875801555c1f8ed1bdd0ba0ca3ab8c8 PDF Text Text Item ID Number 01710 Author Carroll, Ray J. Corporate Author Typescript: Report #4, September 1982 Journal/Book Title Year °ooo Month/Day Color n Number of Images 24 DOSCriptOH NOtOS Report consists of the following sections: The Control Population; Stratification of Battalions; Missing Battalion Exposure Data; An Exposure Index; Sampling Strategies Monday, June 11, 2001 Page 1711 of 1793 �Report f4 1 R.J. Carroll, Ph.D. i September 1982 This report consists of the following sections: T he C o n t ro 1 P opu 1 at ion; A general discussion of some of the i s s u e s to be kept in mind when deciding upon the control (non-exposed) population; S t r a t i f i cat ion of Bat t a 1 i ons; To reduce the cluster effects investigated in Report # 3 , I suggest that battalions first be stratified into similar groups and then sampled. Miss ing Battalion Exposure Data; If many battalion records are not usable for determining exposures, we have to consider some alternatives and their potential effects. C* An Exposure JEndex; I conclude that a priori construction of a ) single, one dimensional exposure index is probably not feasible and I is not even necessary. Multidimensinal indices will be more L appropriate. Sampling Strategies; ! I propose a framework for using multidimensional exposure indices to design the sample. Alternatives are considered and analyzed. �Because there are many unknowns relating to the conducting an Agent Orange study, and because my function ii an advisory one, this report cannot be considered a protocol. I believe this report raises many issues which should be resolved either prior to or by means of the anticipated pilot study. Of course, I continue to be very willing to work with and advise the VA, outside epidemiologists and the survey f i r m picked to develop the final protocol and sample. �Cntro. '• It is! crucial that any 'study of the health consequences of exposure to Agent'.Orange include an .adequate control population of non-exposed troops. The i : ;; ' II'1 \ i \ K \ \ •','-,• ji || •' |iS i .ji | I : I ;i|! •• ' ! !' control or non-exposed population must be so similar to the exposed population I ,1 . , - ' II I ll I r •lii i I** ! t h a t ! we can be sure. ^that differences in nealth status are due to exposure to Agent Orange and not due '^txvdif Eerencesin the make-up of the exposed and non-exposed populations. Conversely, the !non-exposed population must be such that if no detrimental health consequences are discovered due to exposure, vre ; • ;• :; ilj- . ~lj! .|i'ilHMjj:ii|'i| h1 : ' •'.': ip'must be sure that this result is not due to having the non-exposed group less i i . l . l l l i ; ! i . . . * vx healthy than ; anticipated.•' " • •. '; :'I'; , '!• ' I'USl'Vl'ii •' ! '.'• j ; ! r . i l l ' . » i ' | . | Ji; I ' 3 ill : '• ' : ' . ! ' :*&!'•! i l Ifor "' jexample, consider '\' the' situation '!\[,of false negatives, i.e., : : ]| 4j - j ^ . j, ,' i ; i -,i^ h i i • I. ' : soldiers 'reported as not exposed actually had heavy exposure. some My reports #1 ! and ji^'/_jL._,. ^QrQ b^^ <^ 'tj-ieijassunption that;'those listed as heavily exposed • -r^iS't^a c t u a y iiwere • so exposed) make clear ere : ;; , i - ' "ji MilJIi^fl!!! " : i; ! false negative rate is high, j ' lh !' i ; any d e t r m e n t a l h e a l t h"" e f e c t s of Agent Orange will tend to be diluted. We " could thus . : i'M'ili ''I ' !. l!| . ! . j ';i| i j 1 . '/ • • ! : . . I- \'< i ;. I ' •' •' I ' :' ' ' that Agent Orange exposure is not harmful to the ground : •!'.'!i ii, - ' J : '. ' ;! , ! ]; )| !i • ! ; |• ; troops when in fact it is . * : ! : . • conclude .: < I l i that Iif ithe ' : ' I , .• •; ' !I ••, , On the other hand, suppose the non-exposed control group would be expected to be in better health than the exposed group,'perhaps because they had higher . . • • • • • J ' i !l!! ! '' ' / i • ' •' ' ! ' . : ,» ' :: 'I ' . : - ' '! ' I • ;! I . ii, j socio-economic status or were much less :i involved in the field. : : '' i 'i| ' I \' '• ! In this ' instance,".,we will tend to exaggerate the effects of Agent Orange exposure. We • ' ;. I i I •! '. : i. ::i j. •'!.!•! |l ' I i ; !i|i: : ! .'' : could thus conclude that Agent'Orange exposure is harmful to the ground troops : . ;: >\ ' • :;::'!•! i II , i when in fact it is not. What are the characteristics of an adequate control group? First, it �should ideally consist of ,combat troops in Vietnam who were extremely to have,been exposed to any potentially harmful chemicals, ifgongil^, the control groups should be selected to be as similar as possible to the exposed : ! ' ' ' ' group on a set of variables (confounders) relating to health status, e.g., : { ' • ' ' : : i " rji || ! i ': f ! previous disease, MOS, socio-economic status, etc. i I I i I , ' "I !•! I ' . •' : ! : . " : l ; l j : ! II What happens if such a control group cannot be found? Of course, one can '••\{ try substitution of another^ group, e.g., ; I ' ' : ' combat troops who were stationed in Korea at -the same tine or combat troops whose'unit was readied for deployment in Vietnam but sent ei sewhere; Whether such substitution will suffice is not clear to me. : • •« '. i ''::il'|i "I ' . . . One :'group I have not mentioned is non-combat troops who were stationed in * ! ' * • • , 11 : 'i; *• 1 Vietnam. • JFor example, non-combat troops'might take the place of the i • : :i .;| | . :,, i ! : '",' i i'i i . : ' ! !' l i '• . i j 1 ' l 'I 1 \ '• ' non-exposed Vietnam combat ; troops if the latter were not found feasible to ' r'i:li ; ''':!i|'- L :l j ! / • • : • ! • : ) i ,..![ I -. ill ' f i identify.«: There are obvipus confounding problems in that the non-combat troops are likely to have higher socio-economic status, although appropriate ; !;;:T!i H-irNiij'ii; -\\ -• ' '? ' ' stratification can help. Further, a comparison between exposed combat troops and non-exposed non-combat troops mixes up'''effects: is it Agent Orange or merely combat itself (or both) that is to blame? Even if an adequate' control group of non-exposed combat troops is found to be available, it has been suggested that a third group also be established and surveyed: non-exposed non-combat troops in Vietnam. I do not see such a strategy as particularly appropriate if the goal of the VA study is to study the effects of Agent Orange. It may be nice to have three groups and try to simultaneously study the combat experience, but I worry that such a strategy �! ii I will lover the statistical power of what T have been led to believe i,§ i, 1 ,! * ; • • ; • ' ' j' ! pressing ^question: : '. • • ' ' .} ' i effects of Agent Orange. A three group study '; i-'i i . ; - I • !•!• i 'I combat, non-exposed combat, non-exposed non-combat, all in Vietnam) does not seem to me to be appropriate, given limited resources. .;' i �St.rajyjlcay.on of .Battalions ; ; ' i ! . : One eay to reduce the e f f e c t s of sartpling battalions (Report #3, the two-stage cluster sample effect) is by maans of adequate stratification. For example, suppose that ve have identified all those battalions in a group named "Likely Heavily Exposed." I am not necessarily advocating at. this time that I ! such identification be done, but for the moment suppose this is the line of attack decided upon. Having made this identification, it is probably foolhardy merely toirandomly sample a fixed number of "Likely Heavily Exposed" battalions, because these battalions may still be very heterogeneous. . ;j ,; . ,! What needs to bo done is to further classify the "Likely Heavily Exposed" battalions i , M into smaller but more homogeneous subgroups that may be important for types of | exposure, ! areas of operation, degree of actual combat, etc. I am obviously in no position to designate these subgroups or strata, as such stratification ought to be'done by someone more knowledgeable than I about the Vietnam era. ..! Having formed these strata of battalions, we might then randomly select a few battalions from each strata. Such a scheme will tend to be more efficient than the alternative of no stratification, and this improvement in efficiency can be very jlarge. ; • We may also decide to sample from a group of battalions designated "Unlikely Exposed." Stratification can also be done here. It would appear to be sensible but may not be possible to make the strata identical for the , i "Likely Heavily Exposed" and "Unlikely Exposed" groups. ii ! ;| | "•' : :1 'I' Further stratification of individuals within battalions will also be ! : j reduce the effects of confounders. . : :. desirable to �, A point that should also be addressed in the pilot study is this: even in those battalions with adequate exposure data, are the data sufficient for every soldier? Individual exposures may be missing (company records lost, for example) ; \ if i so we have obvious important'difficulties, whieh. are mgre in £he • I • framework of classical sampling theory, i n �There are really two uses for exposure indices, and I think it is vital to keep the distinction in mind. The first use is for choosing the sample. The second use is in analyzing the sample; in this case there is the flexibility to produce a.I number of indices and try to relate them to health status (while keeping in mind, of course, the multiple testing problem). This second use of exposure indices will not really concern me at this time. The key ques't'ion then is the desirability of developing a single exposure index to be used as a vital component of the sampling design. To answer this question one has to return to the: purpose of the study. Stating the study's purpose is not easy for me bo do, but suppose it can be reduced to "Do combat troops in Vietnam who had a heavy exposure to Agent Orange now have poorer health than ;those troops not exposed to Agent Orange?" If this reasonably captures the purpose of the VA study, then we must ask if a single exposure index will adequately distinguish between "heavily exposed" and "non-exposed" troops. upon by all In particular, this index must be agreed the major interested parties before embarking on the sampling, or else seven years hence we might read "VA's Agent Orange Study: were the 'heavily exposed' really exposed at all?". In my discussions with Mr. Levois, I have bcome concerned that there simply is not enough good information available to construct a single �10 ^xposure index on which to base tho'VA study. I think alternatives ihQuM be explored and their consequences studied (see the section of this report "Sampling Strategies"). It might he feasible to convene a panel of physicians and other 'experts from outside the VA with a charge to develop a concensus exposure index which is both medically and politically sound. A d i f f e r e n t conceptual framework for using exposure indices in designing the sample should prove moire f r u i t f u l . The basic idea is to construct a multidimensional index which measures various facets of exposure, and then base the sample on this multidimensional index. I know of one exanple of a study for which I served as a consultant and which used a multidimensional approach quite successfully. The SEN 1C study of the Centers for Disease Control (Dr. Robert Haley, P r i n c i p a l Investigator) was designed to see if programs for surveillance and control of nosocomial (hospital acquired) infections were at all successful. A panel of experts was convened and helped develop a two-dimensional index treasuring the two aspects surveillance and control, and this index was used to choose the sample by stratification. SENIC was also politically sensitive as well as difficult scientifically, and it might serve as a potential guide for the VA (see the American Journal of Epidemiology, May 1980). To give some idea of how a miltidimensional index might be used in the VA study, suppose that "exposure" consists of two conceptual facets, "Aborted Mission Exposures" and "Usual Mission Exposures." A panel of experts is convened and develops an "Aborted Mission Exposure" index (AME index) and a "Usual Mission Exposure" index (UME index). This panel also designates levels of the AME and UMfi indices which are called Tow, Medium and High, forming a �11 matrix such as in Figure #1. Depending on the goals of the Study, tte V& §©Uld then sample from the cells of: Figure #1. For example, all sampling oould be done from the upper left and lower right corners (Low-low versus High-High); such sampling would be sensitive to detecting the effects of Agent Orange. It is my belief that a p r i o r i construction of a muItidimens ional exposure index such as outlined above is more feasible than constructing a single index and nay be very useful in designing the study. N.B. I will later deal with the distinction between low-itBdium~high and a "dose response" type relationship using indices. �In this section 1 will propose and study a framework for a sampling plan which is based upon the idea of a two-dimensional exposure index. To keep the framework simple, I will pretend that simple random sampling is possible, thus ignoring for the moment the more complex issues of stratification, battalion clustering, misclassification and varying exposure levels of individuals within battalions (the first two of which I have previously discussed, although not in this c o n t e x t ) . A f t e r this framework lias been thoroughly studied, I think we will be in a position to confront more of the complex issues. In particular, Report #5 will focus on misclassification as it relates to sampling strategies. For the moment, I want to try to make the basic framework clear and find out if it meets the needs of the VA. As in the section on exposure indices, I will assume that it is possible to roughly conceptualize exposure as either "Aborted Mission Exposure" (AME) or "Usual Mission Exposure" (UME); this is for me just a working hypothesis and should really be explicated by those more knowledgeable than I about Agent Orange. Having conceptualized exposure, in this way, I see a panel of experts (including some from outside the VA) as developing indices ranging from 0-100 which measure AME and UME?;. I then see this panel as developing groups based on the exposure indices, say row-Medium-High for each. Thus, I am envisioning that each soldier can be categorized into low-Medium-High on both the AME and UME indices. This leads to something like Figure #1. (N.B.: I do not know if such a construction can actually be done). To a fairly .large extent, this is the basic scheme used in the SENIC study mentioned in the section on exposure indices. Already, however, many questions �arise: Ql: IS IT NECESSARY TO CATEGORIZE THE EXPOSURE INDICES? Not really, although such categorization is convenient and fairly standard, 2_2: HOW ARE THE CATEGORIES TO BE CHOSEN FOR PICKING THE SAMPLE? Basically, one would hope on medical and not statistical grounds, think it index. I do not would be useful to define IDW AME as the 33rd percentile of the AME Rather, IDW MK should be medically meaningful. Q3: IS IS NECESSARY TO HAVE EXACTLY THREE GROUPS IOW~MRDIUM-HIGH FOR EACH INDEX? No. In fact, one might well want more. SENIC used four, but I use three to make subsequent calculations more transparent. Q£: CAN WE EVER USE THE AME AND UME ' INDICES THEMSELVES AND NOT JUST THE CATEGORIES? Yes, especially in the analysis. One might veil want to develop a "dose-response" relationship based on the indices. ^ PJl: CAN WE USE MORE THAN TWO INDICES? Yes, but I think much more than two indices v«ould become unwieldy. Let us now suppose that the framework of Figure #1 has been accepted (actually, it will inevitably be modified and include various stratifications). How should the sample be picked? purpose of the study. To a major extent, this depends on the Whctt I will now do is consider a few sampling strategies, and then discuss the purpose for •which they are ideal, as well as their drawbacks. S_trateg_y__J 1 (Sample only from the IOW AME-IDW UME and HIGH AME-HIGH UME cells of Figure #2). This strategy is ideal for the purpose of determining whether or not high Agent Orange exposures in ground combat troops are harmful to future health, �14 when compared to an appropriate control group. This strategy will have the highest statistical power for such a comparison and will probably result in the lowest number of misclassifications (see Report #5 to follow). While i Strategy #1 maximizes the statistical power of a comparison between no and heavy exposure, it is not a very good method for estimating a i "dose-response" relationship between exposure indices and health status. In i addition, Strategy #1 w i l l tell us nothing about which of AME or UME is the most harmful, if indeed the High-High exposure group has worse health than the Low-Low exposure group. These may not be much of a drawback, but the principal investigators of the study ought to be the ones making the substantive decision as to the purpose of the study. Strategy ttl is tine mat-hod of choice for the specific question of whether high Agent; Orange exposure is harmful relative to low exposure. It is not an acceptable method if one instead wants to know how risk depends on exposure indices, especially for the middle range of exposure. Strategy #2 (Sample from all the cells of Figure #2, either equally or proportional to size). This strategy has a more general purpose from Strategy #1, being oriented to estimating the relationship between health status and exposure indices, especially for differing amounts of exposure. Further, it will help identify whether an AME is more harmful than a UME. My understanding of the need to compare Low-Low versus High-High exposures and do the best study possible of the effect of Agent Orange suggests that Strategy #2 will not be appropriate. �Strategy #3 (Sample from all cells, but over-sample the WPW~Ti>W and HIGH-HIGH cells). This is a compromise between the two earlier strategies. One way to think of this strategy is to consider it as a three group study, the groups being Low-Low , High-High and others. Such a view is probably misleading. I prefer to think of Strategy #3 as a way of comparing low-low and High-High exposures while at the same time enabling us to get some sort of "dose-response" relationship between exposure indices and health status and some understanding of the relative importance of AME and UME. There being no free lunches, we get both high versus low exposure comparisons and dose-response simultaneously with lower efficiency and higher misclassification errors; the key question is how much efficiency do we lose? It is my understanding that the high versus low exposure comparison is the } most important one, so I will take the view of asking how much can w* sample from outside, the Low-Low and High-High cells of Figure #1 (to get a "dose-response" relationship) before losing a significant anount of statistical power for comparing High-High and low-Low cells. In a later report I will address the problem of additional ndsclassifications caused by using Strategy #3. Suppose that the VA can afford to obtain the health status of a total of N* individuals. We observe N _ L from the Low-Low exposure cell, N™. from the High-High exposure cell, leaving us with N* - N - N = N« to be chosen from the other seven cells. A reasonable strategy for this example but one which may have to be modified is to apportion the N^ - N.^. - N^, = N O observations equally in the remaining seven cells. This is illustrated in ' Figure #2. �An example will help to illustrate the potential problems with sampling from cells other than the LOW-LOW cell and the HIGH-HIGH cell. Suppose that p = Pr (disease in the IOW-IOW cell) = .005 , and the relative risks for disease in all nine cells are as given in Figure #3. Note that I am trying to detect a doubling in risk when the soldier has HIGH UME-HIGH AME ( r Hn = 2 . 0 ) , so that in this example the disease rate for the HIGH-HIGH cell is P^ HH " .01. The other relative risks I have chosen arbitrarily but conservatively. For example, the relative risk for HIGH-HIGHMEDIUM AME is taken in this example as only 1.4, maaning that the dinease rate for this cell is .007 and, in 1,000 soldiers, we only expect the HIGH-MEDIUM cell to have two more incidents of disease than the IOW-IOW cell (versus five more in conparing HIGH-HIGH and IOW-IOW). Further, suppose we agree to follow the UCLA protocol and i n s i s t a p r i o r i that an acceptable probability for concluding that Agent Orange exposure is harmful when it really is not harmful is a = .01. Finally, suppose WG are able to observe the health status of N* - 12,000 individuals. Thus, if we follow Strategy ttl and sample N^ = 6,000 f r o m the Low-Low cell and I. = 6 , 0 0 0 from the High-High cell, the statistical power for detecting the hypothesized doubling of relative risk is 81% (see Report # 2 ) . On the other hand, suppose wa decide ISL^ = 3,900, Nffil = 3,900 and N Q -- 4,200, so we take 600 observations in each of the remaining cells. Then, if we simply compare the disease rate in the Low-Low cell to that of the High-High cell, the statistical power drops to 61%. I will call Anjy.yjs^s_#l_ the simple comparison of the disease rates in the Low-Low and High-High cells. A more complex Analysis #2 compares the disease rates of the combined Low-Low, low-Medium, Medium-Low cells to those in the High-High, Medium-High, High Medium cells; the Appendix gives the technical �17 details.' For Analysis |fl, we see that Strategy #1 had power 81% whil§ St #3 with 600 observations in each of the seven outside cells had power 61%; the corresponding results for Analysis #2 ace 81% and 58%. This drop in power of 20% (Analysis #1) or 23% (Analysis #2) is a serious one and illustrates the following basic fact: Even when there are no misclassification errors, there are choices of p and the table of relative risks for which too extensive sampling from outside the Low-Low and High-High cells causes a serious loss of statistical power in the main coirparison between very low and very high exposures. i / i In the previous example, wa had p = Pr {disease in Low-Low cell) = .005. If we next try p = .01 but keep the same table of relative risks (see Figure #4) then wa get Table #1 M§l.Y?.i§_J.________________^SnELiBa^trategy^J__________Power (p=.005) 1 1 81% 1 3 61% 2 1 81% 2 3 58% (p=.01) I 1 1 99% \ i 1 3 91% i 2 2 i 1 99%! i 3 89% �The drop from 99% statistical power to 89% - 91% power caused! by ^a * ' •• i " " outside the Low-Low and High-High cells when p = .0.1 is still somewhat discouraging, although much less severe than that encountered when p « ,005. This illustrates a second basic fact: For the same table oE relative risks, the loss of statistical pomr due to sampling outside the Low-low and High-High cells becomes greater as the disease probability in the low-Low cell becomes smaller. We next will vary the particular version of Sampling Strategy #3. In the previous version, we took a total of N Q == 4,200 observations outside the Low-Low and High-High cells. In this next strategy, we will take half as much 2j too ( N Q = ir^-eD), thus allocating 300 rather than 600 observations to each of the other seven cells. Note that this second Sampling Strategy #3 will be much less informative about dose-response. The power results are as follows: Table #2 (Second) Sampling JPower Analysis^ (p=.005) 1 1 81% 1 3 73% 2 1 81% 2 3 71% 1 1 99% 1 3 97% 2 1 99% 2 3 96% (p=.01) �These losses of power are much .less dramatic, illustrating the following £ For small disease probabilities in the low-Low cell, sampling outside the Low-Low and High-High cells can be done without much loss of statistical pover if the other seven cells are lightly sampled. From the preceeding analysis as well as many others I have done, I can make some f a i r l y definite conclusions even if I ignore the rnisclassification problem. Using a stringent Type I error rate, a = .01 as in the UCIA protocol and using a conservative table of relative risks, it appears that for the rarer diseases (p s .005) , the ability to detect a doubling of risk going from the ! , \\ Low-Low cell to the High-High cell can be conpromised if the middle range of risk is oversampled. Thus, iC rarer diseases are of major interest (such as p = . 0 0 5 ) , only relatively few observations (less than 20%), if any, should be taken from the middle range of risk. For more common diseases (p ^ .02) or if the goal is to discover a tripling of relative risk (r TirT = 3.0, not 2.0 as tin h e r e t o f o r e ) , then some sampling from the middle range of exposure will entail considerably less potential loss. For diseases which are not so rare (e.g., p ~ .01), no definite conclusion can be made at this point, because I have not yet illustrated the effect of mis classification. It is useful to repeat that only sampling the Low-Low and H i g h - H i g hi cells gives the highest statistical power and i the lowest , , , ; misclassification rate. 1 i �') ] .1 ' j ;[ Additional Remarks There are still many questions that need to be answered. Among these are the following (1) What are the available^ control groups 'of non-exposed troops? ( 2 ) Will, battalions of generally high exposure have troops who were definitely not exposed, or are these false negatives? ( 3 ) W h a t variables can be used for battalion and individual s t rat i f icat ion ? *^*-( (4) What is the extent of missing battalion exposure data? ( 5 ) What features of exposure need to be considered in construction of a multidimensional exposure index? (6) Can a multidimensional exposure index be constructed? (7) For what alternatives (p eind r Strategies) should wa be designing the study? (8) of the section on Sampling What are the difficulties with follow-up to look at health status? (9) How will the power calculations change when the effects of misclassification and battalion cluster sampling are also considered? (10) Can wo estimate the misclassification error rates? �The choice a = .01 made by UCLA is a conservative one ( a » .05 is more ! ;| .i :, ! : ill usual) but still a good idea in setting sample: sizes and discussing effects of • i : ' I different analysis and design strategies. : '" ; ' I : ! ' ill In this appendix, I am assuming that : -i all disease probabilities are fairly low and that, unrealistically, there are i • • 'I ' • '"'I J 'i|! • no misclassi.ficat.ions; the latter case will be dealt with in a future report. For Analysis #1 of the section on sanpling strategies, statistical power is computed as in Report #2, i.e., by treating 2 Arcsin /pTT as normally i ' ^ 1 distributed with mean 2 Arcsin V/pTr and variance , 1/NT1.. lib .Luj For Analysis #2, I am going to coirpare the weighted disease rate : W IlPl,r, ••'• WIM%M j against the i disease rate + W MLPML ,•: W HH%fl where N(L) = W ML = N ML ML/N(L) - NHH/N(H) ' Assuming that the true disease probabilities are all relatively small, it turns • : i1 ' j b out that to a f i r s t approximation, which is sufficient for my purposes of illustrating effects on power of different sampling strategies, 2 Arcsinvp(L) is normally distributed with mean 2 Arcsin /p(L) and variance 1/N(L); a similar .___„ I ,• / ~s^<. < result holds for 2 Arcsin Jp(H). �: • • • I that the approximate :statistical power for 'Analysis ' i1 !: ! This means #2 is Ut A more complex analysis based on additive or log-linear additive tables could also have been considered, but I do not believe the necessary additivity should ! ' ' be assuired when making these important power calculations. i If we assumed a less conservative configuration of relative risks by i '• i changing! Figure 4 to the following (reading across rows) 1.0, 1.1, 1.25, 1.1, 1.25, 1.5, 1 . 7 5 / 1 . 2 5 , 1.75, 2.0, then the power figures in Table #1 would become 81%, 61%, 81%, 66%, 99%, 91%, 99%, 94%, while those in Table #2 would become 81%, 7 3 % , 81%, 75%, 9 9 % , 9 7 % , 9 9 % , 9 7 % . if the interest was in i detecting tripling of relative risk ( r „ = 3 . 0 ) , all sampling strategies gave power of over 99%. the analysis and �.Figure in. '• : ~ " "| ' Jf-> ; C )NCf ,NSUS AMI; TNDIiX '".I i :'!i ABORTED M f S S f O N S ) i i LOW Medium High Low Concensus IJMH Index (Usual Medium Missions) High i These troops had h i g h UMli exposure. ; These troops have high cxpo- surc on botli indices., �1 !;S.i DIST 'Figure #2; OF OBSERVATIONS FOR A TOTAL SAMPLIi OF SIZE CATEGORICAL IN ).!• LOW Medium V7 Low High V7 UMC Categorized Medium NQ/7 NQ/7 Index High N Q/7 'V7 IIH 0 = N* - NLL N For example, if we can take 12,000 observations total, we might take N ( f = 3,900 - 3,900 Ny = 4,200 . �Figure //3 ISSION tiXPOSURE INDEX" -"A !!.f! Low !. Medium, P Low r [,M High LII "Usual Mission lixposurc Medium MII Index" • " KM r. . = relative risk of disease (relative to the Low-Low 13 cell) for soldiers in r i., column jg, i.e., 1 . . 1 -' ™ Pr(disease, row i, column j) '"' '" ~" *" " "* — —. —™ —- I'rfdiscasc, Low on both indices) p ~ Pr(disease, Low on both i n d i c e s ) �Figure #4 A SPK [Fit CXAMPLIi 01' RELATIVE RISKS p '- Pr {disease in the Low-Low cell} . ; - i i ]: Ij! i ' !!: a =j| Type I , error (probability of finding ;.',';:!!-Agent''Orange a health "hazard when it •' ! i''"\"really1 is not) ' "" •:• ' •.; i ' .j Sriii ; ;; r .,j = ' r e l a t i v e risk for disease in the ' 11U j!l I i gKj-High cell J I ., I !! RIiLATIVJ: RISKS ' . Low Low UME Mcdi urn 'LL ML : High Medium ..0 LI I .1 High For an illustration, take r 1 .3 r IIM = 1 4 = 3. .1 .2 � 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. 063 Folder The folder containing the original item. 1710 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/. Creator An entity primarily responsible for making the resource Carroll, Ray J. Title A name given to the resource Typescript: Report #4, September 1982 Subject The topic of the resource Agent Orange Exposure Study study protocol ground troops exposure cohort study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/b6bc3272a58d6a256d3f246c32c91aee.pdf 87025a8766067d6d9a0c0a93b0c773ea PDF Text Text °1709 Item ID Number Author Corporate Author Report/Article TltlB Typescript: Status Report for Development of Exposure Cohorts To Be Used in VA Epidemiology Study, Draft, September 29,1982 Journal/Book Title Year 000 ° Month/Day Color 0 Number of Images 6 Desorlpton Notes Monday, June 11, 2001 Page 1710 of 1793 �September 29, 1982 Status Report for Development of Exposure Cohorts to be used in VA Epidemiology Study Of prime importance to the proposed VA Epidemiology Study of the effect of exposure to Agent Orange on the health status of Vietnam Veterans will be a clear, thorough statement of the process used to identify study subjects. The procedure used by the Array Agent Orange Task Force (AAOTF) has been designed to make this selection feasible, efficient and as unbiased as possible. To minimize the very large number of records to be reviewed and data to be entered a multistage process is used. The basic selection or "sampling" frame will be the military unit which will be characterized as "exposed" or "not exposed". Individuals will then be identified from selected units and further studied to assure their experience is correctly represented by the unit's classification. The pilot phase of this study will determine whether adequate cohorts of exposed and non-exposed can be defined. To accomplish this an audit trail will be maintained of all units identified, the proportion selected and the basis for exclusion of the non-selected units. Thus the pilot phase will also perfect the selection process to be used during the full study. For a meaningful comparison of outcomes among subjects "exposed" and "unexposed" to Agent Orange, cohorts should be as similar as possible in other aspects of the Vietnam experience. Thus, units should be stratified to reduce heterogenicity. It should be sufficient to select exposed and unexposed units from the same branch of service, in a similar type of unit, and operating in the same Corps. If possible, similar terrain would be a useful forth criteria. In the first step in finding unexposed units, candidates are selected from among those that were operating beyond the reasonable maximum drift for Ranchhand spraying. It will be necessary to evaluate other potential exposures (e.g. perimeter spraying, dumps, etc.) for a final classification into exposed or unexposed unit. �The availability, completeness and accuracy of the various record systems is unknown at present and this will be studied during the pilot phase. However, the procedural guidelines or selection algorithm used by the AAQTF is to a large part based on assumptions about these records. This increases the likelihood the AAQTF staff will encounter unexpected variations. If this requires AAQTF staff to make eligibility decisions on units or individuals not anticipated in the algorithm, a brief ancedotal record of the nature of the problem and the convention adapted to solve it is most important. This will facilitate an evaluation of the cohort selection process for potential selection bias during the pilot phase and will determine the feasibility of the proposed epidemiological study. The following procedures are for selection of Vietnam service subjects with high and low likelihood of exposure to Agent Orange or other herbicides while in Vietnam and do not include the selection of a non-Vietnam cohort. The outlined steps have been included in the guidelines supplied by the AAQTF and should provide a list of subjects who can be recruited for a pilot study. The documentation requirements provide information to evaluate the feasibility of idenitfying adequate cohorts for the full-scale study. I. Prepare maps of Vietnam with Ranchhand spray tracks (both defoliation and crop) for the period 1967-1968 (this has already been provided). II. Define geographically homogeneous areas with known ground troop activity with and without Ranchhand tracks during the 2-year period. This is the first step and defines areas which might be selected. The rationale for choosing selected areas must include areas with and without Ranchhand tracks, similar location (and if possible terrain), and the presence of ground troops. If possible, all four Corps of RVN should be represented. 2. �III. a. Obtain appropriate station lists for each area selected. b. Determine potentially eligible units Which were operating in selected areas during the 2-year period. All units identified from station lists are potentially eligible. IV. Develop "enhanced" HERBS Tape by including information from herbicide applications not identified from Ranchhand missions. This will include perimeter sprayings, road clearings, helicopter missions, aborts and other identifiable applications. It should not be necessary to do more than those areas previously identified in Step II above and it will be necessary to include the entire period of study (2 years) for this effort in those areas to be studied. V. Determine availability and completeness of all eligible units' records of movement. VI. a. Select units to be recorded as to daily movements. This is a crucial step, and documentation of how and why a unit (or units) were selected should include the eligibility of all units which were potentially eligible and might have been selected. The potential for the selection of adequate numbers of units (or subjects) for a full scale study may depend on records' completeness and availability at this step. Obviously only a small subsample (one or two areas perhaps) can be done in the Pilot Study. b. Record to machine readable format the daily location coordinates for selected units during the entire 2-year period. A necessary part of evaluating the reliability of exposure estimates will be the assessment of completeness of the available records for herbicide applications. Documentation of completeness will assist in this effort. VII. a. Match daily location coordinates of selected units via computer with the "enhanced" HERBS Tapes to determine the number and type 3. �of "hits" sustained by each eligible unit between January 1, 1967 and December 31, 1968. Each "hit" will be defined and recorded according to all of the following four parameters for each unit: i) Date ii) Type of application (Ranchhand, Helicopter, Aborts, Ground incidents) iii) Time in days since application (1st, 2nd, 3rd, 4th) iv) Distance from application site in km. (0-.49, .5-.99, 1.0-1.99, 2.0-4.99) (if spray or abort). b. List the "hits" for each unit matched and the weighted score. An index or weighting scheme is attached. While it may not be possible to quantify exposure with certainty, at least a weighting rule has been developed to classify high and low exposure units. The last category in both time and distance parameters is intended to provide a buffer between exposure and nonexposure. A "hit" in either of these categories will not constitute an additional exposure in otherwise exposed units but will exclude otherwise unexposed units from being considered as unexposed. Thus, "low likelihood of exposure" units will not have been within 5 kms. within 60 days of a Ranchhand application target if the pilot shows this is feasible. VIII. Select high and low likelihood of exposure units. This step will depend on selection via the weighting system developed for the previous step as well as the results from the Pilot Study. The rationale for selection will be indicated and should include other criteria for the type of units chosen as well as herbicide exposure. Depending on the results of Step VII, it may be possible to identify "low likelihood of exposure" units with no "hits" of any kind. If this is the case, then a specific weighting system to discriminate between high and low exposed units will not be crucial for 4. �selection. "High likelihood of exposure" units can then be selected from among those units with at least several "hits" of various kinds (exclusive of buffer zone hits). IX. a. Obtain daily morning reports for the period January 1, 1967 to December 13, 1968, for each unit selected in the previous step, b. Track individuals through service with their respective units during this time period and record presence on "hit" days for their unit. Some criteria for eligibility will have to be developed and documented, including the minimum number and quality of individual's "hit" days, reliability of records for identification of individuals during this step, and their minimum length of assignment to the unit. Initial criteria are appended. X. a. Obtain service records for selected individuals and verify service information obtained via morning reports, b. Record personal information from service record and include location of medical record for each selected individual. At this step, it may be desirable to retain the individual's identification as either exposed or unexposed during the period covered in these procedures. This should enable closer matching of the two groups based on individual characteristics obtained from the service record. It will be necessary to identify multiple tours of duty in Vietnam and perhaps other criteria pertaining to military service which will influence his final selection into either of the Vietnam service cohorts. Documentation of these inclusions and exclusions will be necessary in order to assess bias in the final selection. 5. �Although most of these procedures are included in the guidelines supplied by the MOTF, they are detailed here to identify the steps and type of documentation which will be useful in the evaluation of the exposure index and determination of the feasibility of selecting exposed and unexposed cohorts for the full scale study. Systematic departures from the procedures outlined here which may be necessary for the inclusion of subjects from each of the four branches of military service should be included in the documentation. 6. � 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. 063 Folder The folder containing the original item. 1709 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/. Title A name given to the resource Typescript: Status Report for Development of Exposure Cohorts To Be Used in VA Epidemiology Study, Draft, September 29, 1982 Subject The topic of the resource Agent Orange Exposure Study study protocol cohort study HERBS database ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/69edfa3a08278eb993611d67f0834391.pdf 223b96049409a21d68c5c2d176df1505 PDF Text Text Item ID Number 01707 Author Corporate Author RepOTt/ArtlGlB TItiB "Typescript: Agent Orange Epidemiologic Study Decision Paper, September 15,1982 Journal/Book Title Year °00° Month/Day Color Number of Images D e Descplpton Notes Monday, June 11, 2001 Page 1708 of 1793 �September 15, 1982 AGENT ORANGE EPIDEMIOLOGIC STUDY DECISION PAPER The U.C.L.A.-submitted epidemiology protocol for the study mandated by Public law 96-151, has undergone extensive review and revisions. Many of the reviewers, especially members of the White House Agent Orange Working Group (AOWG), have recommended widening the focus of the study. Although U.C.L.A. has prepared a protocol for strictly an Agent Orange study (high and low likelihood of exposure cohorts of Vietnam veterans) it has been suggested that a third cohort consisting of Vietnam era veterans (no Vietnam service) be included to assess the Vietnam experience. Some reviewers (e.g., members of the VA Committee on Health Related Effects of Herbicides) have proposed that the VA conduct a study of the Vietnam experience only. If the results are positive, further efforts to elucidate a cause and effect would be possible but not required. It is apparent that at least three courses of action are available to the Agency. 1. Conduct an Agent Orange study based on the current version of the U.C.L.A. protocol. This study would compare the health outcomes of two groups of Vietnam veterans, those having a high or low likelihood of exposure to herbicides. 2. Conduct an Agent Orange study but add a cohort of veterans that did not serve in Vietnam. This study would involve a modified version of the present protocol and would compare health outcomes between two cohorts that served in Vietnam and one that did not serve in Vietnam. 3. Conduct a study of the Vietnam experience. This study would compare the health of veterans that served in Vietnam with veterans that served in other parts of the world during the same period. �The options and major arguments for selecting or rejecting each are as follows: OPTION I. SHOULD THE VETERANS ADMINISTRATION CONDUCT A PURELY AGENT ORANGE STUDY (TWO COHORTS IN VIETNAM)? PRO VIEW o An Agent Orange study was mandated by Public Law 96-151. o The U.C.L.A. protocol has been designed around a two cohort study of Vietnam veterans (high and low likelihood of exposure). o A purely Agent Orange Study provides a single issue focus for the VA since the perception is that this is the major concern of Vietnam veterans, o The AOWG Science Panel Chairman on November 18, 1981, testified before the Senate Veterans Affairs Committee that an Agent Orange study can be conducted, o An AOWG Science Panel Subcommittee on "Cohort Selection and Exposure to Herbicides" indicates that cohorts for an Agent Orange Study can likely, and objectively, be found by using DOD records, o Many scientists are genuinely concerned over potential low dose effects of dioxin and Agent Orange, o A two cohort study may be conducted more rapidly and at less cost than a three cohort study, o A pilot study can be conducted rapidly because fewer changes in the present protocol will be needed and fewer subjects need to be recruited and examined. ANTI VIEW o Public Law 97-72 permits the VA to expand the focus of study. o There are scientific problems associated with conducting a purely Agent Orange study involving two cohorts: �- there are no clearly defined long-term health outcomes attributable to TCDD and Agent Orange exposure (the Air Force Health Study consisted of a very extensive health examination because of this). - the military records necessary to determine objective assessment of exposure in Vietnam were never designed for use in an epidemiologic study; thus, there is a significant potential for misclassifying individuals as to whether they were or were not truly exposed to herbicides. - there is no scientific way to equate modes of exposure in Vietnam (e.g., how many perimeter exposures equal a RANCH HAND exposure?) o Errors resulting from misclassification of individuals within the high and low exposure cohorts will seriously impair the useability of data, o DOD and selected members of AOWG are against conducting only an Agent Orange study. They believe that the scope of the study is too limited, o A pure Agent Orange study may resolve the concerns raised by some Vietnam veterans. However, if the study shows no effect attributable to Agent Orange, many veterans will say it was something else in Vietnam that impacted their health. OPTION II. SHOUED THE VETERANS ADMINISTRATION CONDUCT AN A3ENT ORANGE STUDY WITH AN ADDITIONAL COHORT TO ASSESS VIETNAM EXPERIENCE (THREE COHORT STUDY)? PRO VIEW o This study may permit some answers as to the impact of the Vietnam experience on the health of Vietnam veterans. 3. �o Some government agencies (e.g., DOD) and some Congressmen favor a broader study, o Many veteran groups may favor a broader study so long as the VA adequately addresses Agent Orange, o The Air Force Health Study adequately addresses the health issues associated with heavy exposure to herbicides including Agent Orange, but it does not address the impact of the Vietnam experience, o DOD believes that a non-Vietnam cohort can be selected without great difficulty. ANTE VIEW o The protocol for an Agent Orange Study may not be suitable for a Vietnam experience study. Hence, to study both issues with the same protocol may compromise the quality of both studies, o The present U.C.L.A. protocol would require significant modification in order to incorporate selection criteria for a third cohort. Moreover, such modifications of the present protocol may have public as well as scientific repercussions, o The identification of a third cohort will require an expanded records search effort by DOD. This will require additional time and resources, o To conduct a study of 18,000 individuals (three cohorts @ 6,000) will require significantly more resources than will be required for the study of two cohorts. OPTION III. SHOULD THE VETERANS ADMINISTRATION CONDUCT A SEPARATE EPIDEMIOLOGIC STUDY TO ASSESS THE VIETNAM EXPERIENCE? PRO VIEW o If the Veterans Administration is concerned with the broader issue of the health of the Vietnam veteran, it must then be equally concerned over the effects of infectious diseases, illicit drugs, combat stress, herbicides, insecticides, etc. 4. �o This study may provide the Agency with significant information to aid in the development of a compensation policy. ANTI VIEW o This Option will require the preparation of an entirely new protocol. It will require time for protocol preparation and scientific review, o Because this study must be "something for all"f i.e., it must be widely focused in terms of the potential health outcomes, it will require indepth physical, biochemical, and psychological studies of the subjects and by necessity may be more costly and take longer to conduct. RECOMMENDATIONS: 1. It is recommended that the VA conduct an Agent Orange Study as designed by the U.C.L.A. Protocol. This study will focus on the health of two cohorts of veterans who served in Vietnam. The issue is whether there is a long-term health effect as a consequence of exposure to Agent Orange. The impact of the total Vietnam experience must be viewed as a separate issue. The U.C.L.A. protocol has received extensive review, and needs only to be slightly revised (Sections on Cohort Selection and Biostatistical Methods expanded) prior to the award of a contract for the conduct of a pilot study. The present protocol can be completed within the next 6 weeks and Request For Proposals (RFPs) for the conduct of a pilot study can be initiated within 60 days. APPROVE/DISAPPROVE DONALD L. CUSTIS, M.D. Chief Medical Director APPROVE/DISAPPROVE ROBERT P. NIMMO Administrator 5. �2. In order to address the issue of the Vietnam experience, the Agency will conduct the Vietnam Veteran Mortality Studies and the Vietnam Veteran Identical Twin Study. These studies, in association with the CDC Birth Defects Study and Australian Studies will bear on the issue of the Vietnam experience. However, the Agency should explore the feasibility of a separate epidemiologic study to address the impact of the Vietnam experience. APPROVE/DISAPPROVE APPROVE/DISAPPROVE DONALD L. CUSTIS, M.D.. ROBERT P. NIMMO Chief Medical Director Administrator 6. � 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. 063 Folder The folder containing the original item. 1707 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/. Title A name given to the resource Typescript: Agent Orange Epidemiologic Study Decision Paper, September 15, 1982 Subject The topic of the resource Agent Orange Exposure Study study protocol study criticism cohort study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/246090beba16f13cc364c7bc2573d0c6.pdf 73aec24bf8ddb11775b6f8a52e56d419 PDF Text Text °1704 Item ID Number Author Corporate Author ROpOrt/ArtlGlO Title Typescript: Status Report for Development of Exposure Cohorts to be Used in VA Epidemiology Study, Septembers, 1982 Journal/Book Title Year 000 ° Month/Day Color Number of Images D e Descrlpton Notes Monday, June 11, 2001 Page 1705 of 1793 �September 3, 1982 Status Report for Development of Exposure Cohorts to be used in VA Epidemiology Study Of prime importance to the interpretation of any results from the proposed VA Epidemiology Study of Vietnam Veterans' Health Status which might be associated with exposure to Agent Orange or the "Vietnam Experience" will be an understanding of the process used to identify potential study subjects. It is under- stood that the procedures developed by the Army Agent Orange Task Force have been designed to make this selection both feasible and as efficient as possible, and entail a multistage process to minimize the very large number of records to be reviewed and data to be entered. Since one of the major goals of the pilot phase of this study is to determine whether and how meaningful exposure cohorts can be selected, the successful completion of the pilot phase will entail an evaluation of the cohort selection process as well as other aspects of the proposed study. In order to accomplish this goal, it is necessary that there be maintained a thorough documentation of the disposition of those units identified but not selected and the basis for inclusion or exclusion of all units considered. Furthermore, the procedures used to select study subjects for the pilot phase should mimic those to be used during the full study. In order to insure a meaningful comparison of health outcomes among subjects "exposed" and "unekposed" to Agent Orange, cohorts should be similar with respect to other aspects of their Vietnam experience. Thus, the units from which individual subjects will be chosen should be "matched" in some ways. Sufficient similarity should be maintained if exposed and unexposed units are selected from the same branch of service and a similar type of unit, and operating in the same Corps �-2- area, preferably in similar terrain. A more clearcut difference in exposure can best be assumed if unexposed units are selected from among those that were operating beyond the reasonable maximum drift for Ranchhand spraying and/or following a sufficient length of time for degradation of sprayed herbicide. Of course, it will be necessary to evaluate other potential exposures (e.g., perimeter spraying, dumps, etc.) for a final classification of both exposed and unexposed units. Since the availability, completeness and accuracy of the various record systems is unknown at present, this will have to be determined during the pilot phase. This both prevents a more precise procedural guideline at this time and forces the AAOTF staff to make eligibility decisions at several steps in the cohort selection process. The detailed documentation of the basis for decision at each step in the selection of units from which subjects will be chosen will facilitate an evaluation of the cohort selection process for potential selection bias during the pilot phase and will be necessary for determining the feasibility of the proposed epidemiological study. The following procedures are for selection of Vietnam service subjects with high and low likelihood of exposure to Agent Orange or other herbicides while in « Vietnam and do not include the selection of a non-Vietnam cohort. The outlined steps have been included in the guidelines supplied by the AAOTF and should provide a list of subjects who can be recruited for a pilot study. The documentation requirements should provide information for an evaluation of the feasibility of identifying adequate cohorts to be used in a large full-scale morbidity study among Vietnam Veterans. �-3- I. Prepare a map of Vietnam with Ranchhand spray tracks indicated for the period 1967-1968 (this has already been provided). II. Define geographically homogeneous areas with both Ranchhand tracks and ground troop activity during the 2-year period. This step should include areas which might be selected and the rationale for choosing selected areas, including the presence or absence of ground troops as determined from Command Post locations (if available). III. a. Obtain appropriate station lists for each area selected. b. Determine potentially eligible units which were operating throughout all or most of the 2-year period in the same area. This step should include documentation of all units identified from station lists determined to be potentially eligible and how this determination was made. IV. Determine availability and completeness of all eligible units' records of movement. V. a. Select units to be recorded as to daily movements. This is a crucial step, and documentation of how and why a unit (or units) were selected should include the eligibility of all units which were potentially eligible and might have been selected. The potential for the selection of adequate numbers of units (or subjects) for a full scale study may depend on records' completeness and availability at this step. b. Record to machine readable format the daily location coordinates for selected units during the entire 2-year period. VI. Develop "enhanced" HERBS Tape by including information from herbicide applications not identified from Ranchhand missions. This will include perimeter sprayings, road clearings, helicopter missions, aborts and other identifiable applications. It should not be necessary to do more than those areas previously identified in Step II above and it will be necessary to include the entire period of study (2 years) for this effort in those areas to be studied. �-4- A necessary part of evaluating the reliability of exposure estimates will be the assessment of errors in and completeness of the available records for herbicide applications. Documentation of the type and number of errors and changes as well as some method for assessing completeness will assist in this effort. VII. a. Match daily location coordinates of selected units via computer with the "enhanced" Herbs tapes to determine the number and type of "hits" sustained by each eligible unit between January 1, 1967 and December 31, 1968. Each "hit" will be defined and recorded according to all of the following 5 parameters for each unit: i) Type of 11) Date, iii) Type of Ground, iv) Time in Agent (Orange, White, Blue, Unknown). application (Ranchhand, Helicopter, Aborts, incidents) days since application (1st, 2nd, 3rd, 4th through 60th). v) Distance from application site in km. (0-.49, .5-.99, 1.0-1.99, 2.0-4.99). b. List the "hits" for each unit matched. An index or weighting scheme will have to be developed (not by AAOTF) by this time. While it may not be possible to satisfactorily quantify exposure, at least a weighing rule must be developed in order to identify high and low exposure units. The last category in both time and distance parameters is intended to provide a buffer between exposure and non-exposure. A "hit" in either of these categories will not constitute an additional exposure in otherwise exposed units but will exclude otherwise unexposed units from being considered as unexposed. Thus, "low likelihood of exposure" units will not have been within 5 kms. within 60 days of a Ranchhand application target. VIII. Select high and low likelihood of exposure units. This step should be documented, but will depend on selection via the weighting system developed for the previous step. The rationale for selection should be indicated and should Include other criteria for the type of units chosen as well �-5- as herbicide exposure. Depending on the results of Step VII, it may be possible to identify "low likelihood of exposure" units with no "hits" of any kind. If this is the case, then a specific weighting system to discriminate between high and low exposed units will not be crucial for selection. "High likelihood of exposure" units can then be selected from among those units with at least several "hits" of various kinds (exclusive of buffer zone hits). IX. a. Obtain daily morning reports for the period January 1, 1967 to December 31, 1968, for each unit selected in the previous step. b. Track individuals through service with their respective units during this time period and record presence on "hit" days for their unit. Some criteria for eligibility will have to be developed and documented, including the minimum number and quality of individuals' "hit" days, reliability of records for identification of individuals during this step, and their minimum length of assignment to the unit. X. a. Obtain service records for selected individuals and verify service information obtained via morning reports. b. Record personal information from service record and include location of medical record for each selected individual. At this step, it may be desirable to retain the individual's identification as either exposed or unexposed during the period covered in these procedures. This should enable closer matching of the two groups based on individual characteristics obtained from the service record. It will be necessary to identify multiple tours of duty in Vietnam and perhaps other criteria pertaining to military service which will influence his final selection into either of the Vietnam service cohorts. Documentation of these inclusions and exclusions will be necessary in order to assess bias in the final selection. �-6- AHhough most of these procedures are included in the guidelines supplied by the AAOTF, they are detailed here to identify the steps and type of documentation which will be useful in the evaluation of the exposure index and determination of the feasibility of selecting exposed and unexposed cohorts for the full scale study. Systematic departures from the procedures outlined here which may be necessary for the inclusion of subjects from each of the four branches of military service should be included in the documentation. At this time, there are several important issues pertaining to the selection of potential study subjects for the full study which require further development before final specification can be agreed upon. Although results from the pilot study selection itself will contribute to the resolution of some issues, there are several problems which can be and must be resolved before completion of this phase. These are listed below and will be developed more completely in separate reports: 1) The inclusion and details for selection of a third non-Vietnam service cohort and (perhaps) a fourth "Vietnam Experience" cohort. 2) The individual criteria from personal service (201) records which may determine eligibility for the study (specificallyinduction status and the possibility of multiple tours). 3) The exact time and distance parameters to be recorded for herbicide "hits" with the possibility of widening the buffer zone. 4) A weighting scheme for the quality of exposure to specific herbicides for the various time, distance and type of application "hits." 5) The advantages and feasibility of extending the 2-year (19671968) period currently provided for in the present procedures. � 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. 063 Folder The folder containing the original item. 1704 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/. Title A name given to the resource Typescript: Status Report for Development of Exposure Cohorts to be Used in VA Epidemiology Study, September 3, 1982 Subject The topic of the resource Agent Orange Exposure Study HERBS database study protocol cohort study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/483fd208b62083787a20464319147e8b.pdf 7c75b68136b6c55bec5f8bf9928be6ef PDF Text Text °1702 Item ID Number Author Corporate Author Report/Article TltlO Requests for Proposals: Military Records Search and Vietnam Service Indicator for Patient Treatment File (PTF) JOUmel/BOOk TltlO Commerce Business Daily Year 1982 Month/Day August 20 Color n Number of Images 2 DOSCrlDtOn NOtOS RFPs were Monday, June 11, 2001 forwarded to Alvin L. Young by D. Layne Page 1703 of 1793 �REASON FOR RBPUHBNCK Lj ron VOUR AS RiequiisTso FICBS NOTE AND HSTURN pKff CONVKMtATIOH NftCfORAMY ACTION TSmSStS tlttNATUNK OAT4 Ja****^"'' T f . KXTf AUO ' OC VA C USED. '^m1^^^^^^^^^'^^1^'^^^^^,^^^^^-1^^^'**^' ** •;'" -"WBWBWHW^w -ssiflw, -wf*»"iewM8*" 5!s'^^*=" �age 2 yujts/sj rr _„ .:M)a? wiling, new underground secondary wiring, Iwlwlrts ami bracket amis comieo to ptwide a system complete and raxiv («use—lFB N62467-80-B-4539- -This ocuremcnl limited to small business only. BOD 17 Sep 92. (229) WCC PO Box 54307 HAS Memphis Mlliinjton, TN 38054 * A - SOlOfl MAXIMUM MISSION SOFTWARE OPERATIONRFP 5 -70340/ '06. Applied ^search Corporation. See role 46. (229) NASA. (Mini SIMM Flight Center (GSFC), SreenbeltRoad, GwenMtt, MO 20771. Attn: Richard Swunson, Coda 208 * .1 -- SPACE SHUTTLE SYSTEMS INTEGRATION for the George C. Marstoll iace Flight Center, Marshall Space Flight Center, AL 35812. RFP 8-1-3-SA-10000-44B 'ill be released a/a 11 Aug 82, on a sole source basis to Rockwell Intentional Corporaon, Space Operations/Integration & Sattelite Systems Division, Downey, CA 90241. '291 Procurement Office, George C. Marshall Space Flight Center, Marshall Space FHtttt Center, Al 356! 2 A - AIR QUALITY TECHNOLOGY, RESEARCH, DEVELOPMENT. AND BE10KSTRATION FOR LAND MANAGEMENT NEEOS-HFP-28-K2.257 due 8 See ,2-..fst $60,000 to $75,000. (229) USDA Forest Service, Rooky Mountain Forts! and Range Experiment Station, 240 West Prospect Street, Fort Collins, CO 60528 Attn: R. E. Caruerry, Contracting Officer COMMERCE BUSINESS DAILY PjlilMiijittJllJIJM^jyj0[|^ .81&T; ,~_^__ output electrically paralleled with the normal power source. Sol HGW-^-it-tmi (229) David Taylor Naval Ship Research and Development Center, BetHesd», MO 20084, Attn: Paul McFartana $«di 332, 202/227-4219 A - FIVE MONTHS R&O EFFORT ENTITLED "ASSESSMENT FOR HARD Ma OF ARM"- * RFP is available—Interested parties staid contact the prime contractor direct—See note 27—Contact: W.R. Injnrji, Attn: DRSKI-IIBB, 20V876-8851 irtad to negotiate with Cas, Its 555 Sparkrnan Dr, Suite 1002, Huntsville, H 35805. (226) U.S. Army Missile Command, Directorate of Procurement I Produc\ Hen, Xedstone Arsenal, AL 35198 "A* 1*4 H Export and Consultant Services 5l*»*WI Dwelopnwt is negotiating with It © ® H - TRANSCRIPTION SERVICES--IF8 BATF-82-24-Closing date August 31, 1982—Contact Jackie Barber 202/566-7421. (228) Bureau of Alcohol, Tobacco and Firearms, Procurement Section Room 2212, 1200 Pennsylvania Avenue, N.W., Washington, OC 28226. OH- OALL ISLAND lOgOlNG AND TRANSPORTATION PLAN-Provide for transportation and timber harvest alternatives foi the Dall Island planning area (16,600 acres) Including appropriate written reports and oral presentation The project area is SmeAH^'^V Virginia Ce«trKSM| *«tf»ftl. WWhlngton. OC 28301, Attn: Ca located spprox 60 miles west of Ketchikan in southeastern Alaska. Estimated start work WafQ * A - "DEVELOPMENT OF A SYNTHETIC FUEL RECIPROCATING CHARGE date-September 27, 1982. All work to be completed and final reports submitted by till close-of-business January 30,1984. lechnical competence will be evaluated using the PUMP".—trie Department ol fnergy, Pittsburgh Energy Technology Center is negotiating following criteria with relative importance as shown: I Experience in logging engineering Mb Ingcrsoll Rand Company, Princeton, NJ on a sole source basis. See note 46.1229) including: demonstrated ability to locate landings on aerial photos, maps, and on the U.S. Department o Energy, Pittsburgh Energy Technology Center, PO Box 10840, Pitteburi, PA 15238. Attn: tequtettla* and Assistance Si. Alaska and the Pacific northwest to solutions of timtw harvest problems. Maximum Division MS Sufl-L Michael Hogan ra to the contracting ** HP *t «te 9/13/82. iZfl) points possible: 25.2. Working knowledge ef forest road planning and location, including I*««Y R. Ryan, ts*««« Mfestr, Offlte »f Procurement Supply »»>; Vi^HSwWai Office, Room 756, * A - DIRECTIONAL FREQUENCY ANALYSIS AND RECORDING (illFAR) check Hie locations accurately on Dw ground. 25 pts. 3, DMonstrated ability to perform ViiM^ujfcfttfKMll*" SONOBUOY MODIFICATION.- NAVAIRSYSCGMIIQ is conducting a procurement for the design, development, fabrication and test of a high density field (HDF) DIFAR sonobuoy. A from field or map profile surveys in live, standing, and multispan skyline modes. 25 pts. O H - EN9IHEC«lia KMlll«eWtn«TlOH SUPPORT SERVICE confidential request for info will be mailed mid-August 1982 to qualified AnVSSQ-538 4. Has suitable computational systems and software to analyie skyline payload, at any NAVALiSSSlmK*•**^ •"*• ^- 83 terrain point, analysis tor live, standing, running, and multispan skylines, dhe software SUPPORT OFinstallation M srtwwrTi^^s contract period of one year FOf My, assembly, Diego, CA. RFP N00244-82-it-l» HUH jfl Sop 82. Sonja Norman, Code 203, band acoustic detection capabilities. Us urgency of the requirement makes it impracticmode.) 15 pts. 5. Knowledge of silvicuftural systems and the logging systems required for /35-3168. satisfactory accomplishment. 10 pts. 6. Technical proposal. 40 pts. 7. Price proposal. 10 * H - CONTRACTOR SSWMf HI MKtt OF PROOUCTING SIMULA! ties. This restriction does net preclude any independent assessments and/or ideas M ' pts. Persons/firms desiring consideration should request RFP RIO §2-89, tai submit all other sources deem pertinent to this development. For further into and a copy ol the RFI, data they wish considered relative to Mr qualificatijms. No collect calls, please. AND TRAINING award anticipgiad to Cs*ato Sciences Corporation, San Dteg under the option provisions of rj^^t ITO5-8J-0-0304, T. Holt, Code 201.4, contact Mr Aaron Sclinitzer (AIR-21532G), 202/592-7656-WWAI8 synopsis no. 258Responses due September 14,1982. (228) 235-3283. (228) ' 82. (229) 9. R. Kerst, Contracting Qffleer, USDA, Forest Service, PO Box 1621. , Naval Supply Center, K1 |, feti* Or., San Oiago, CA 82132 NAVAIRSYSCOMHO. (A-02E1), Washington, DC 20301 lutteau, AK 99102, 907/SW-7480 * A - IONS ENDURANCE FLIGHT DEMONSTRATION OF ALBATROSS 3008.. Negotiations are being conducted en a selected source basis with leading Systems, Inc., Hacienda Heights, CA. See note 46. (229) Dufense Supply Service-Washington, Roam 10245, The Pentagon, Washington, D.C. 20310 (Altn: Pat Barton) H - MILITARY RECORDS SEARCH: Ihe purpose is to provide informtien to VA wtnch can be used to study the cause of death among Vietnam era veterans and sited light on their health status. RFP 101-31-82 available by 8/13/82, upon written request to the Contracting Oflicer. RFP due date is 9/13/82. (229) Jeffrey R. Ryan. Contracting Offlcsr, Office ol Procurement Md Supply (03B), Veterans Administration Central Office, Room 758, J10 * A - LEASE OF A SUM, STIRLING CYCLE ENGINE GENERATOR SET FOR Vermont Ave., N.W., Washington, OC 20420 TESTS AND EVALUATION PURPOSES.—Hegotiations will be conducted with Stirling Power Systems Corporation, 7101 Jackson Road, Aim Ate, Ml. DRDK-FtA/Jenkins/703/ OH- ONSITE KEYPUNCH/VERIFY SOURCE DATA to diskettes or 80 column 664-539! • cards services for a period of one year with two cm-year options—COCO equipment and * A - DEVELOPMENT OF A LIGHTWEIGHT LINEAR ALTERNATOR FOR A facilities—Indefinite delivery requirements lypt of contract- Place ot performance NavFREE-PISTON STIRLING ENGINE.- -negotiations will be conducted with Mechanical al Air Station, Jacksonville, Fl—IFB H00207-82-B-003(WX»313 Sept 1982. (229) technology, Inc., Latham, NY. OROMC-PEA/Jenkins/703/664-5391. (229) Contracting Officer, U.S. Naval Air Station, JscksonvlH*, FL 32212. U.S. Army Mobility Equipment Research and Development Command. Attn: E. Kelly Bllven, 984/772-5132 Procurement and Production Directorate, Fort Betnoir, VA 22060 * A - NUCIEAR MAGNETIC RESONANCE GYROSCOPE, unsolicited Negolia lions are to be conducted with Hughes Research Laboratories, Malibu, CA 90265. See note 40. * A -- TEXTURE ANALYSIS FOR ROCK-TYPES DISCRMINATION SUMMARY OF PHASE I RESULTS AND PROPOSED TASKS FOR PHASE II EFFORTS. Unsolicited. Negotiations are to be conducted with SusqiKtena Resources and Environment, Inc., Johnson City NY 13/90. See note 40. (229) Department of the Air Force Mr Force Office of Scientific Research (AFSC) Boiling AFB, DC 20332 A - ENGINEERINS RESEARCH STUDY TO OEVEOP THEORETICAL ANALYSES OF A FLOATING UNINTERRUPTIBLE POWER SUPPLY CONCEPT. In the system a static convertei is connected between the battery and electrical load. The converter, which is capable of bi-directional ppwrtfflow. Floats on-line with its A.C. Ihe Commeree Business Daily iUSPS %6-360) is (rtilisW daily, wept Saturdays. Sundays ant holidays, lor J175 a year (1st Class mlin|) or 1100 a y(s» (2nd Class mailing) by the I) S Gowmenl Prating Ollice. Washington. DC 20402 Second Class postage paid at Washington. DC. Ml additional mailing ottices POSTMASTER: Sent address changes to Sueerint«W of Dumts. IIS Govwimral Printing Mien. Washniton. DC 20402. will entne mailing label Iron lasl issue received UK Secretary ol Commerce has **™i«cd that the publication ot this periodical is necessaiy K the transaction ol the public business tequned by to ol this Decartmnt Use of Ms lor printing this periodical lw, Ira waved by the Director ol Ibt Mce ol taagtmat and Budget ttuough 31 July .1% FURNISHING PLAN MILITARY BAIIHIACKS-Parl II—The contractor shall prepare a consolidated research report on furnishings in military barracks, guard rooms, and hospitals for the period 1981 to 1940. Hie est level ot oHart approx six person months - RFP CMlQO-lflV-317 issued o/a August 31,1982. Proposals doe a/a September 20, 1982—all requests shall be submitted in writing only—Envelope requesting RFP should be marked "Request for CX -1100-IIW-317"—Availability of the sol is limited and will be made upon first received first served basis until supply is exhausted. (228) National Park Service, Harpers Fen; Center, Harpers firry, WV 2542S, Attn: Annette Akshlro. Contrast Specialist, 104/135-0371, e«t, 6238 - VIETNAM SERVICE INDICATSR FOH PAYSENT TREATMENT HIE (PTF). The contractor will be required to review a selected simple of military molds and codify certain inlormaticn relating to Vietnam service. RFP 191-29-82 available up* written request. RFP due date 9/13/82. (229) leifray R. Ryan, Contracting Officer, Office of Procurement and Supply (038), Veterans Administration Central Office, Room 7S8,910 Vermont Ave„ N.W., Washington, BC 26420 " ~ "ff11OF CU«KW ASStTS *W INCOME. Reports are to del the ability of debtors to the U.S. Granrw« to repay these debts. IF8 is GAPS. Procurements Branch, U,S, («,^ Accounting flflica. 441 N.W., Washington, DC 20541 tarn Jt45, Attn: Ruth Ann Marls tract Specialist. •«"«»• H -- SUPPORT SERVICES FOR WO mil BUSINESS CONFERENCE held in Washington, OC and Phoenix, H m 1953 n« support services include-onj publicizing, conducting and reportim on the conferences Iho area of consider) restricted to small business firms. RFP 82-122 isstisl approx August 27, 19 below. (229) national Sdense Foundation, Division of Grants & Contracts, Street. N.W. Room 031, Washliitton, OC 20550, Attn: Mary L. H (RFP82-1X2) H -- PHOT PROIECT ON PROGRAM EVALUATION STUDIES. Pro| 82-1. Ihe wk will consist of facilitating three rwfci evaluation protects. The under consideration are: 1. Literature proem-iitmry magazines and small category. 2. Design arts program-design itanstration category. 3, Office < Chairman for Management-analysis and use ol final descriptive reports from gran deadline for proposals is October 12,1982. The level of effort will be the equ '""'""' .---.- j.|,,ta Ritseirch Division, National Endowment for the Ara. 2401E Washington, DC M336, Attn: Program Sol 82-1. O H -TESTINGTHE HALON 1331VFIRE EXTINGUISHING SYST rious areas, site R. PA. Sol DA!A08-82-d-0453 available o/a 24 Aug 82. Bid i 21 Sep 82. Packages available on a first come, first serve basis. (229) Procurement Division. Hindquarters, Fort RlUihle. MO 211 Theresa M. Wing, 301/8T8-S705 , H - SMALL GRANT FOR SUPPORT OF RESEARCH IN THE AREA USE REGULATORY REFORMS. HUD will award two grants ol up te $?5.t individuals or organizations to research and document successful land and H - COMPUTER SERVICES FOR THE U.S. ARMS CONTROL AND DISARMAMENT AGENCY (FY1982)--F!OA AC20X702, Me hfent CHO, Order 002-Control Data Corp., 6003 Executive Blvd., Reckville, MD 20852. (229) The U.S. Arms Control and Disarmament Agency, 2Ui and Virginia Avwue, N.W, Washington, DC 20451. in housing cost production. The small grants are to stimulate and support the ( written materials that will be useful to local officials, coalitions of civic or * H - PERFORM SUPPORT SERVICES INVOLVING THE ENGINEERING of housing. Requests tor application will be available August 12,1982 througi � 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. 063 Folder The folder containing the original item. 1702 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/. Source A related resource from which the described resource is derived Commerce Business Daily Date A point or period of time associated with an event in the lifecycle of the resource August 20 1982 Title A name given to the resource Requests for Proposals: Military Records Search and Vietnam Service Indicator for Patient Treatment File (PTF) Subject The topic of the resource Agent Orange Exposure Study ground troops exposure cohort study ao_seriesIII 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. 063 Folder The folder containing the original item. 1701 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/. Creator An entity primarily responsible for making the resource Fein, George Title A name given to the resource Letter: from George Fein to Barclay M. Shepard, August 16, 1982 Subject The topic of the resource Agent Orange Exposure Study study protocol cohort study study criticism ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/4bb27c46613704dd7a53cea895d5ee84.pdf 237206a28c86f8aa4ffbe86c9a9fd9c5 PDF Text Text Item ID Number Author 01700 Carroll, Ray J. Corporate Author ROpOrt/ArtlGlO Title Typescript: Report #3, The Effect of Sampling Battalions Rather than Individuals, August 1982 Journal/Book Title Year °ooo Montn/Oay Color Number of Images D 5 Descrlpton Notes Monday, June 11, 2001 Page 1701 of 1793 �Mi REPORT n i' THE EFFECT OF SAMPLING'BATTALIONS RATHER THAN INDIVIDUALS j '• j l.J. Carroll • JAugust 1982 In my letter of August; 14jto M.E. LeVois, I raised the possibility that sampling:] from; battalions and thehlisampling individuals could have different statis- i W 1!:V . j - ' . i . iil'i : tical properties from merely taking a simple random sample of individuals. The former method is mentioned by,UCLA in their protocol, but it is clear that they I •""«.' i ' : '! , intend tojuse the latter as a basis for analysis. I think it is important to ' :• I • .". understand the difference between the two and to investigate the effects of this difference. This report is a preliminary analysis of this difference. While the latter method is called simple random sampling (SRS), the former i method might best be called two-stage cluster sampling (TSCS). The two methods j • ' t i are illustrated in Figures tfl-and #2. j •: ' . ; 1 '! : !'' | Suppose there are a total of M battalions and, for simplicity, assume each ! : i ; - /••!.' j .i ' i ! : :ji battalionjhas m individuals. 'Let battalion #i; (i = 1,2,..., M) have disease . i '••'•!• : and ';i rate p. suppose ' ! overall ! disease i •!;; ili-! the: ' I _ • rate is p. Suppose we randomly select ; •Mi 1 ! '-f1! 1: ! J ! i j i i:. N battalions and then select n jindividuals per'.selected battalion. Then the 1 ,\ i estimated probability of disease ,is (for either method) the observed proportion of diseased individuals. If the probabilities;of disease are all :fairly small ; , ! : i i "j!|i : • (say less I than 3% in every battalion), then the variances are approximately ;! I ' • ill ' „ 'Vanance(SRS) = ^ — P • ' ' ! ' mnc". • (1-nvn/MN) — ' ; i (| .. f , i Variance(TSCS) SjiilB/ll)- + ¥ «x j.w.ji'ws* v, & vui~y — . 777 n r i ' i '-S ;' '. I'-i'l',-).^ • ! •' 1 ' ': i ; '*; : i m ' £l ! ' " * • *' ' !' i '''I -i'l :j 't^'ij j i '! ::' L M ^ ^ "1 -*1 "] . : Suppose'there are M = 500 battalions, of whichjwe sample m * 70. ^Suppose that i : ; ) ' i ji'|;ri «j|.i.j, < t i I ; j '! • ' jij' j '.. j' j|| �each battalion has N - f>00 individuals, of whom we sample 100, Pui'the?, §V»ppPse that, the disease probabilities are all less than 3%. Then, to a degree of ap.i proximation, , . Variance (SRS) = (.0118)2p i ' 'iir ' »~ Variance(TSCS) '= [.0001149 + |:0030715p]p . Table #1 compares the ratio of'these two quantities for various values of p. The exact numbers in Table j#l are not particularly critical, especially since any real sampling plan wi.ll include some stratification. Nonetheless, my calculations indicate the following: i| . . .:* . .! (i) As a general strategy, we should sample as many battalions as •i i : .'Mi • •'!• possible, with appropriate stratification, i J •! | i ' (ii) For larger sample:sizcs on the order of 6,000 per group, if the event rates are small the effect of TSCS will not be too great, I , (iii) If the event rates are large, TSCS will be significantly less i •": efficient than SRS, However, in this instance, we will still have acceptable statistical powers (see Report #2). i ., , This^report has notjaddressed certain problems, such as confounders and misi ,1 <, classification. Also, I have;assumed the event rates are fairly homogeneous ' • • ' I i • ; 'ii iii; I ;!!: across battalions; this might\be a questionable assumption, as it is conceivable : :I • j ; '! ; that a few battalions had extremely high exposure and event rates. : • '•!; ii ' ' ••' study must be guided by the' practical nature of the data set. i Further �Figure _//1 SIMPLH RANDOM SAMPLING OF TRN INDIVIDUALS FROM A TOTAL OF FOUR BATTALIONS. ,Battalion #1 Battalion 112 Battalion Battalion #4 -X X X X X X X X X X J, o oo x x Note: © © On average, we will choose someone from every battalion. tee this by taking a stratified sample. ; We could guaran- *^*1 Figure #2 TWO-STAG!- CLUSTHR SAMPLING OF TRN INDIVIDUALS, SELECTING AT RANDOM TWO OF FOUR BATTALIONS. •i,'?! : Battalion til Battalion Battalion ix x x x x O | O Dip O X X X X X o; o bio o i * * * * * •ii * *' * *, * Battalion «f4 © <& © ffl © © © e ® ® Select Battalions #2 and #3 i j j i :•' . Battalion ;';!' Battalion ( x-%.: • ; ;: ,00000 I fo o o o o I U i / / O O O O O • ; :!• ' , ! . ! : •; * * * * * i Note: As opposed to simple random or stratified random sampling, in cluster sampling there is no chance of selecting one or more individuals from i| I every battalion. t ••"i! • 1$ ! : i ii .! ' �TABLH #1 Approximate Value of P .005 ,010 .020 Variance *(TSCS) .95 1.04 1.26 I !| �Appendix, Report #3 Variance (SRS) - -^ —'/—.'„ p v( i _lp ) inn Variance (TSCS) = ' - , * in I ' nm c2 2 , M S/ = -JT V (p.-p)2 1 n-1 .'•, 1 1 ' ' M S 2 = M(¥-lT .? Pi If the event rates arc all small, — 2 ] —2 (p.-p) < j p on average. � 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. 063 Folder The folder containing the original item. 1700 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/. Creator An entity primarily responsible for making the resource Carroll, Ray .J. Title A name given to the resource Typescript: Report #3, The Effect of Sampling Battalions Rather than Individuals, August 1982 Subject The topic of the resource Agent Orange Exposure Study study protocol cohort study ao_seriesIII https://www.nal.usda.gov/exhibits/speccoll/files/original/bef3e3fd9f02a5c3a14d1c89984b77d1.pdf f5fa65935a6165dc90e58cca4ac7f456 PDF Text Text Item ID Number °1699 Author Carroll, Ray J. Corporate Author RflpOrt/ArtlOlO TltlO Typescript: Report #2, False Negatives and their Effect on Power of Statistical Tests of Differences in Disease Rates, August 1982 Journal/Book Title Year 000 ° Month/Day Color Number of Images DOSCrlptOU NOtBS D 4 See ltem 1773 for a letter ^rom Carro"that mentions this report. Monday, June 11, 2001 Page 1700 of 1793 �REPORT 112 FALSE NEGATIVES AND Till:111 EFFECT ON POWER OF STATISTICAL TF.STS OF DIFFERENCES IN DISH ASF. RATES R.J. Carroll i August 1982 In REPORT #1, especially Appendix H3, I calculated statistical power for tests i' ' ' to see if disease .rates are the-same for exposed and non-exposed soldiers. I concluded that, if the disease rate in the non-exposed population was at all substan| • ; "-« •• tial (say 5%), then samples of 6,000 for each group would give very high statistical power for detecting a relative risk due to exposure of 2.0 or more, even if 1 " there was substantial misclassifixation of exposed into the non-exposed group ,i (false negatives). In this report, I will study the effect: of false negatives on statistical! power when the disease rate in the non-exposed population is fairly small. \.i i; 'To review terminology, recall that ! : j! ' ' ;;l . i p = disease probability;for truly exposed • i !• q = disease probability;for reported exposed ~ p. ' i "' i' p '= disease probability,for truly not exposed q '= disease probability for reported not exposed .•' ••• i 1 r I ~ ep,/p- = true relative risk true l*2 . . ( r ; i ' = ^i/^o " observed i iRelative risk OUi? A »•• ' 6 =ifraction of false negatives i N •= sample size for reported exposed N? = sample size for reported not exposed. .'i Before1constructing a test for equality of exposed and not exposed, we need : ! !II two concepts. The level or Type 1 error a of a test is the probability (usually a ; ,• : i I 1 .05 or*! 5%) that we will ' i conclude that theI exposed have different disease proi . ! '1 i ' in babilitiesi than do the.non-exposed, but in reality the two are not different. �The power of a test (V - 3) is the probability we will conclude that the exposed differ from the non-exposed when they really do differ. Generally, we want small Type I error but high power; UCLA chooses a = .01, 1-0 = .95. One point not made clear in the UCLA protocol is that they compute sample size based upon a one-sided test and not the usual and more conservative twoi ; In this particular instance 1 think1what they have done is justifi- sided test. ! ' able; we can hardly expect exposure to agent orange to have improved health status, If we assume equal sample sizes so that N I = N? = N, then based upon UCLA's arcsin transformation, the statistical power of the usual test is given by the formula 1 - <I»(z.. J. ~0t where z - /-=-(2arcsin/ii v ^ •*• - 2arcsin/q.J) £ is the (l-a)100% pcrccntile of a standard normal distribution function J. ~*vJt ' $ and q2 = 6p + (1-<S)P2 (Report #1, Appendix If we want to detect a true relative risk r of at least r. ., we will then : . true detect p . I • 'i need a minimum sample size of at least i M N = : 2(z if we want power of 1-3. Note that this is exactly the formula on page 108 of the UCLA protocal as long as there are no false negatives. Some idea of the effect of i false negatives can be learned from Table #1. Figure #1 shows this effect even i , . more dramatically. Note that if we desire a Type I error of a = .01 and power i !' 1-3 a .95|to detect a true relative risk of 2.0 or more, then we need about 4,500 soldiers inieach group if there are no false negatives; we will need over 7,600 �soldiers in each group if there are 20'-i false negatives. This report shows that for diseases which are fairly rare, the effect of misclassification can be severe. EFFECT OF FALSK NEGATIVES ON CALCULATION OF SAMPLE SIZE a • 1-3 r. . . detect lp~ 2 6 UCLA N Actua 9,124 11,721 15,397 20,828 29,305 4,528 5,815 7,637 10,328 14,527 2 , 230 2,863 3,757 5,078 7,138 .01 .95 2.0 .01 0.0 .1 .2 .3 .4 9,124 9,124 9,1 24 9,124 9,124 4,528 4,528 4,528 4,528 4,528 .01 .95 2.0 .02 0.0 .1 .2 .3 .4 2,230 2,230 2,230 : 2,230 2,230 .01 .95 2.0 .005 0.0 .1 .2 .3 .4 N__ �«4 l or \ \ f r \ u \ --f-O! � 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. 063 Folder The folder containing the original item. 1699 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/. Creator An entity primarily responsible for making the resource Carroll, Ray .J. Title A name given to the resource Typescript: Report #2, False Negatives and their Effect on Power of Statistical Tests of Differences in Disease Rates, August 1982 Subject The topic of the resource Agent Orange Exposure Study study protocol cohort study ao_seriesIII