June 13, 2001
Dr. Grant LeFarge,
Secretary and Member,
New Mexico Board of Medical Examiners
2nd Floor, Lamy Building
491 Old Santa Fe Trail
Santa Fe, NM 87501
RE: The Ethics of Medical Determinations regarding polygraphs
Dear Dr. LeFarge:
Thank you very much for taking time to talk with me today. For you records, let me summarize the history of the specific problematic issue that I believe will soon face physicians who care for patients who are employees at Sandia or Los Alamos National Laboratories.
In early FY2000, the US Department of Energy (DOE) instituted a polygraph counterintelligence "screening" process, mandated by the Congress in the 2000 Defense Authorization Bill. The Polygraph Program was to focus on "national security" questioning, including questions on the use of classified information, contacts for foreign nationals, and attempts at espionage. In the course of implementing the Polygraph Program, the DOE held hearings at each of the National Laboratories purportedly to take comments from scientific staff who were to be subject to the polygraph. At these Hearings, several staff members from each of the Labs, including Dr. Lawrence Larsen and me (two of the three MDs on the scientific staff as distinguished from the Occupational Medicine Department) questioned the routine practice of polygraphers asking subjects for medical information including all extant or past medical conditions and a list of prescription and non-prescription drugs being taken to treat those conditions. An exhaustive review of the medical and scientific literature conducted by Sandia's 26 Senior Scientists (both Dr. Larsen and I also happen to be Senior Scientists) revealed no scientific studies demonstrating any effect of any medications or medical conditions on the outcome of the polygraph (which, in any case is an extremely inaccurate test when used in a screening mode -- see attached articles). We presented this information at the Hearings, which was subsequently ignored by the Department. However, Paul Robinson, Sandia's president, was sufficiently moved by the invasion of privacy (and by Dr. Clevenger's findings -- see attached memo of March 6, 2001) that he demanded that the DOE stop asking medical questions, or he would withdraw Sandia from the polygraph program.
A series of negotiations and meetings began in late March between Sandia and the DOE Office of Counterintelligence (OCI). I attended one of these meetings, chaired by our Senior Vice President, which included the DOE Chief of Polygraphy (Mr. David Renzelman) and the local DOE "testing center" polygrapher (Mr. John Mata). At this session, the polygraphers continued to insist that (a) medications have an "effect on the polygraph" and (b) that these effects were consistent and reproducible. When I asked for references (articles, reviews, or textbook chapters) to substantiate their claims, they could produce nothing. Instead they made glib, illogical statements like: "well, if you're all hyped up don't you think that your pulse will be changed" (my answer: "of course, but is it differentially changed when your subject is actually lying as compared to telling the truth"). Ultimately, Sandia seemed to prevail and the OCI indicated it would change its policy on requiring medical information.
That policy "change" is what you will see in the first attachment: a memo from Mr. Michael J. Waguespack, Director of the Office of Counterintelligence at the DOE and sent to all of the National Laboratories and DOE polygraph "testing centers." At first blush, it appears that OCI has relented in its demands for private medical information. In the third paragraph, Mr. Waguespack states: "The practice of completing a medical consent form has been eliminated". However, as you read on you'll see that DOE has now placed the burden entirely on individual Lab staff -- and their physicians -- to come forward with "medically related" information that could affect the polygraph. How any physician (let alone NON-physician) could know what constitutes such "medically related" information in the absence of any scientific study is, regrettably, not of concern to the Department. Rather, they have imposed a duty on Lab employees and their doctor's to identify such information. Implicit is that there may be adverse action taken against the employee if they fail to reveal "medically relevant" information.
But, my concern here is for physicians who may be consulted by Laboratory employees seeking to fulfill this onerous new responsibility. After thorough searching of the medical and other scientific literature (utilizing a broad search strategy in Science Citation Index and Medline) I was able to locate three articles (attached) that address the diagnostic characteristics (sensitivity, specificity, and positive and negative predictive value) of the polygraph. The information, as expected, demonstrates a complete absence of diagnostic utility when the polygraph is used in the screening mode (as opposed to the criminal investigation mode where the characteristics of the test are somewhat different and the pre-test probability is markedly different). Few physicians will be aware of such data, yet they may soon be deluged with requests from worried patients who face not only the injustice of spurious screening by polygraphy, but also the burden of having to evaluate what are, as yet, the indeterminate possible effects of particular medical conditions on their test outcomes. I believe that the Board would provide a valuable service by apprising physicians of the absence of documented validity of the polygraph as a screening test. While physicians equipped with this information might, on principle, and without prompting, decline to furnish testing approval of recommended waivers for patients having particular medical conditions, it would surely be best for the Board also to advise physicians directly that, given the absence of data establishing that any particular medical condition may confound polygraph test results, they should abstain from supplying patients with declarations that could either clear patients for, or exempt patients from, polygraph testing on the basis of the presence of a medical condition. Such a directive would serve the additional purpose of ensuring that physicians do not risk committing the ethical infraction of responding to unreasonable demands from government authorities, potentially to the detriment of their patients.
Even in the realm of granting security clearances, individuals are entitled to a fair, repeatable, and unbiased process of fact finding and judgement. The Polygraph Program is none of these things, and, in my view, stands to further harm patients as DOE seeks to illicitly obtain validation of the entire process through the co-opting of the medical community. Patients deserve better than this.
Sincerely Yours,
Alan P. Zelicoff, MD
Senior Scientist
Center for National Security and Arms Control
Sandia National Laboratories
Albuquerque, NM 87185-5800
Work phone: (505) 844-8020 Work FAX: (505) 844-0600
Home phone: (505) 255-6908
e-mail: apzelic@sandia.gov
Attachments:
- Memo from DOE Office of Counterintelligence dated 5 May 01 [sic, correct 31 May 01]
- Memo from Dr. Larry Clevenger to Sandia Executive VP Joan Woodard, 6 March 01
- Articles (#3)
- Polygraph. Council on Scientific Affairs, American Medical Association, published in JAMA, Sept 5, 1986 -- Vol 256(9): 1172-1175
- Brett, Allan S. et. al. Predictive Power of the Polygraph: Can the "Lie Detector" really Detect Liars? Lancet, March 8, 1986, I: 544-47
- Steinbrook, R. The polygraph test -- A flawed diagnostic method. NEJM 327(2): 122-23, 1992