EDGEWOOD

Opening

From 1955 to 1975 the Army ran a medical volunteer program at Edgewood Arsenal in Maryland, where service members were exposed to chemical agents and drugs. The substances at the centre of the record include LSD, the incapacitating agent BZ, and nerve agents, among them VX and sarin. The men were volunteers on the forms. What they were told, and what the Army owed them after they went home, is the argument that fills the later paper.

The headcount is not one number. In Vietnam Veterans of America v. Central Intelligence Agency, the government admitted that about 7,800 service personnel were used at Edgewood and that 250 to 400 chemical and biological agents were administered there. The Ninth Circuit in 2015 described about 6,700 subjects and more than 250 agents. A Defense Science Board summary of the National Research Council follow-up counts 6,720 soldiers in a 1958-1975 program of 24 chemical agents. None of the sources explains the gap.

Timeline

26 February 1953Secretary of Defense Charles Wilson sends the memorandum later cases call the Wilson Directive to the secretaries of the Army, Navy and Air Force. It sets conditions for human volunteers in atomic, biological and chemical warfare research. Its most quoted line: the voluntary consent of the human subject is absolutely essential.

30 June 1953Brigadier General John Oakes, Secretary of the General Staff, sends memorandum CS:385 to the Chief Chemical Officer, the Surgeon General and other senior officials. It adds limits on the agents: controllable lethality, no serious chronicity anticipated, effective therapy available, and an adequate background of animal experimentation. It also says medical treatment and hospitalization will be provided for all casualties of the experimentation as required.

1955The Ninth Circuit dates the Cold War wave of this work, nerve agents and intensely psychoactive chemicals, primarily at Edgewood, from this year.

26 March 1962The Army promulgates Army Regulation 70-25. Judge Claudia Wilken's summary-judgment order of 24 July 2013 gives that date and says the regulation was reissued in 1974, and the Ninth Circuit says the reissue left the consent and casualty-care language unchanged.

21 July 1975The Vice Chief of Staff orders an Inspector General inquiry into volunteers in chemical agent research, written by Colonel James R. Taylor and Major William N. Johnson of the Investigations Division.

1975According to the Ninth Circuit, the Army stops large-scale experiments exposing human subjects to chemical agents.

10 March 1976The IG report is dated: Use of Volunteers in Chemical Agent Research, report number DAIG-IN 21-75. A declaration by Caroline Lewis Wolverton, filed in the veterans' case on 15 September 2010, attaches excerpts under that title and date.

8 August 1979Army General Counsel Jill Wine-Volner writes to senior Army officials and the Surgeon General about notifying participants in drug and chemical or biological agent research.

24 September 1979She sends the Director of the Army Staff what she calls broad guidance for a notification program.

25 October 1979Lieutenant General John McGiffert, the Director, tells the staff to identify and notify participants who may face a continuing injury.

1980The Army asks the National Research Council's Committee on Toxicology to study possible chronic or delayed effects in the Edgewood volunteers.

1985The final volume of the NRC's Possible Long-Term Health Effects of Short-Term Exposure to Chemical Agents reports on the test subjects' current health, based on a questionnaire that, the Defense Science Board later says, drew 4,085 replies out of 6,720 people tested.

30 September 1988A new AR 70-25 takes effect, with a "duty to warn" and a sentence authorizing medical care for injury or disease that is a proximate result of participation. By what the Army later called a serious editing mistake, an appendix lists deliberate exposure to chemical and biological warfare agents as exempt.

1990The Army reissues AR 70-25 and corrects the exemption. The Ninth Circuit said in 2015 that this version was still in force.

2009Vietnam Veterans of America (VVA), Swords to Plowshares and individual veterans sue the CIA, the Defense Department, the Army, the VA and named officials: No. C 09-0037 CW, Northern District of California, before Judge Wilken.

30 September 2012Wilken's class certification order recites the government's admission of about 7,800 personnel and 250 to 400 agents.

19 November 2013Wilken enters judgment. The Army loses on the notice claim. The plaintiffs lose on the claim that the Army, rather than the VA, must provide medical care.

30 June 2015The Ninth Circuit decides Nos. 13-17430 and 14-15108. It affirms the notice injunction, holds that AR 70-25 also imposes a medical-care duty, and sends that part of the case back.

Key figures

The volunteers are the subject of the program, not a backdrop. The judgment names Bruce Price, Franklin D. Rochelle, Larry Meirow, Eric P. Muth, David C. Dufrane, Tim Michael Josephs, William Blazinski and Kathryn McMillan-Forrest, along with VVA and Swords to Plowshares. Their pleadings are their account, not findings. Rochelle's complaint says he was at Edgewood from 1 September 1968 to the end of October 1968, was told not to talk about the tests, and was promised follow-up care he says never came. Meirow's says that in October 1972 his company was asked for volunteers at morning formation and told they would be testing military equipment.

James S. Ketchum was an Army psychiatrist who, the complaint says, became Edgewood's first psychiatrist in 1961. His obituary in the September-October 2019 Dartmouth Alumni Magazine says he joined the medical research volunteer program at Edgewood, rose to head of clinical research and chief of behavioral sciences, and retired from the Army in 1976. He died at home in Peoria, Arizona, on 27 May 2019. The obituary notes a 2012 New Yorker profile. He wrote Chemical Warfare: Secrets Almost Forgotten, a personal account of testing incapacitating agents on Army volunteers.

Wine-Volner's two memos are the clearest internal statement that silence was no longer a policy the Army could defend. Wilken tried the suit. Judge William A. Fletcher wrote the Ninth Circuit opinion, with Judge J. Clifford Wallace concurring in part and dissenting in part.

Physical and documentary trail

The record includes the Wilson Directive, CS:385, the four versions of AR 70-25, the IG report, the 1979 memos as the Ninth Circuit sets them out, the NRC series as the Defense Science Board summarized it, two Federal Register notices on volunteer record systems, the pleadings, the 2013 judgment and the 2015 opinion.

The consent language is worth reading slowly. As Wilken's 2013 order quotes the 1962 and 1974 regulation, voluntary consent is absolutely essential. The volunteer must understand enough to make an informed decision, "so far as such knowledge does not compromise the experiment." He will be told as much of the nature, duration, purpose, method and hazards "as will not invalidate the results." The same passage says he will be fully informed of the effects on his health or person that may possibly come from participation. The Army wrote both the promise and the limit into one paragraph.

The IG report's contents page shows what the 1975-76 inquiry thought it was examining: chapters on psychochemicals and LSD, the derivation of authority, selection and screening, informed consent, and chemical warfare experimentation with humans at Edgewood. It also lists field tests at Fort Bragg, Fort McClellan, Fort Benning and Dugway Proving Ground, and an Intelligence Corps LSD section with subsections titled Edgewood Phase, Third Chance and Derby Hat. This report sticks to the arsenal's volunteer program. The approval block names Lieutenant General H. N. Maples, the Inspector General and Auditor General.

Wine-Volner's August memorandum, as the Ninth Circuit quotes it, says the Secretary of the Army had concluded that some notification program was necessary as policy, and that "the legal necessity for a notification program is not open to dispute." She cited a Department of Justice conclusion that another federal agency might well be held to a legal duty to notify drug-testing subjects whose health it had reason to believe might still be affected. Her September memo ordered a review of programs aimed at military, rather than medical, uses of these substances. Participants at risk of continuing injury were to be told of their participation and what was known about the substance, "regardless of whether the individuals were fully informed volunteers at the time." McGiffert's memorandum also told the Surgeon General to keep watching substances whose long-term hazards were still unknown.

The NRC work is easy to over-read. The board says the 1958-1975 tests ran at the Army Chemical Test Center at Edgewood, and that in 1980 the Army asked the committee two questions: were the data good enough to estimate long-term harm, and were the chemicals as tested likely to cause it? Three volumes followed. The first covered anticholinesterase and anticholinergic chemicals; the second, cholinesterase reactivators, psychochemicals, mustard gas and several irritants; the third, in 1985, current health.

The board's reading is neither a clean bill nor a casualty list. On the anticholinesterase agents the panel found no evidence of long-range harm at the doses used, yet its members "were unable to unequivocally rule out" long-term effects in some individuals. On the anticholinergics, it found no firm evidence of long-range harm, said uncontrolled variables made the behavioural data hard to judge, and thought long-term toxicity or delayed effects unlikely on the data it had. Mortality in both groups was not significantly above the US population matched by age and calendar year, and the 1985 questionnaire found no significant long-term effects or clustering of problems separating exposed men from the unexposed or the general population. A review of VA and Army hospital admissions for cancers, mental disorders, and diseases of the nervous system and sense organs showed no significant excess against the unexposed baseline group.

Things you might not know

• The consent rule had an escape hatch. AR 70-25 told the volunteer he would be fully informed of possible effects on his health, and in the same breath told the researcher to hold back anything that would "invalidate the results."
• Chemical-agent research was exempted by a bad edit. The Army's 1990 summary of the change called the 1988 exemption a serious editing error, made while moving a paragraph. The wrong version lasted one revision.
• The written secrecy oaths were not in the files. Plaintiffs said volunteers were sworn to secrecy and threatened with a general court-martial if they talked. Wilken's 2012 class order notes that the defendants could not locate written secrecy oaths from the Second World War or the Cold War.
• The recruiting pitch, as one plaintiff tells it, was equipment and a long weekend. Meirow's complaint says the formation was offered three-day weekends and an extra $2 a day to test military equipment, with details to come on arrival. Rochelle's says his Edgewood pay for travel was $1.50 a day, plus a certificate.
• The follow-up database was not only soldiers. A Federal Register notice of 11 December 1981, quoted by the Ninth Circuit, describes records on volunteers who were military members, federal civilian employees and state prisoners, in Army tests of potential chemical agents and antidotes from the early 1950s until the program ended in 1975.

Competing explanations

The Army's account is that these were volunteers, consent was required, the agents were supposed to have a known treatment, and doctors were on hand during the tests. In the Ninth Circuit the government accepted that the 1988 medical-care sentence required care for a proximate injury, but argued the duty ended with the experiment, and that the notice duty did not reach volunteers from before the 1988 text. The court rejected both limits. Nothing in the sentence confines care to the days a volunteer is on the ward, it said, and the 1990 correction showed the Army meant the rule to cover people deliberately exposed to chemical and biological agents.

The veterans' account is that "volunteer" described the signature, not the knowledge: they were not told what the substances or risks were, secrecy kept them from asking a doctor for years, and notification came late and thin. Rochelle's complaint is one version: a gas-chamber test, a gap in his memory, record codes he says were never explained in plain language, and no follow-up before he was sent to Vietnam. The 2013 judgment did not try those histories, but it did hold that an ongoing notice duty exists.

The NRC follow-up is a third account, about epidemiology rather than consent: no clear excess of long-term illness or death in the groups it could study, and no certificate that the anticholinesterase exposures were harmless. A questionnaire that cannot separate exposed from unexposed men does not say what a named soldier was given on a named day, or whether his later illness came from it.

Later life of the case

On its face the suit is broader than Edgewood: the CIA is a defendant and the complaint sweeps in other testing. The duties the courts enforced are Army duties under AR 70-25, and the 2013 notice class covers service members who were test subjects in any Army program that exposed humans to a chemical or biological substance to study its effects.

On notice the veterans won: the Army must tell class members of new information that may affect their well-being as it becomes available, and an injunction "shall issue."

The Ninth Circuit agreed on notice, including for volunteers from the older experiments, and reopened care. Wilken had been wrong to refuse an injunction on the ground that the VA was already treating some of the same people, and that part went back to her.

Notice in practice was already partial. Wilken's 2013 order records that the Defense Department had given the VA the names of 16,645 Cold War-era test subjects, a list wider than Edgewood, and the VA had written to about 3,300 for whom it had an address. The letters did not name the agents. The government cited time, a database still changing, and a wish not to alarm people about unlikely effects. That is notice of a kind, not notice of what a man was given.

Finding

The program is documented. From the mid-1950s to 1975 the Army exposed volunteers at Edgewood to chemical agents and drugs including LSD, BZ, VX and sarin. Its 1962 regulation required informed consent and, in the same paragraph, let experimenters withhold whatever would spoil the test. In 1979 the Army's General Counsel wrote that notification was legally necessary. The courts enforced an ongoing notice duty in 2013, and in 2015 held that the regulation also requires medical care for a proximate injury after the test is over, sending that injunction back to be written. The best description of the consent is the regulation's own split sentence, not a later slogan about it.

What still will not resolve: how many people the program included (somewhere between 6,700 and 7,800, depending on who is counting), what each was given, and which later illnesses that exposure caused. The follow-up studies found no clear excess of illness and gave no certificate of safety either. Many of the volunteers have died, and a letter that never names the substance does not answer the question the regulation itself raised.