University of Arizona's Violations of Federal Regulations in Dr. Kay's Case

Home ] Write to Arizona Legislators ]

University of Arizona's Violations of Federal Regulations in Dr. Kay's Case

A letter from Frederick K. Goodwin, M.D. to Chris Pascal, J.D., Director of the Office of Research Integrity (ORI)

April 28, 1998

Chris Pascal, J.D., Director
Office of Research Integrity
Suite 700
5515 Security Lane
Rockville, MD 20852

Dear Chris:

As we recently discussed, I'm writing to provide you with additional information relating to the University of Arizona's investigation of Marguerite M. B. Kay, M.D., Regents' Professor of Microbiology and Immunology and Medicine at the University of Arizona Health Sciences Center.

Dr. Kay, a National Institute of Aging grantee and an internationally-renowned investigator of molecular and immunological mechanisms of human aging, is facing serious misconduct charges by the University. I was retained by Dr. Kay's counsel to assist in her defense, and in reviewing the evidence relating to the University's charges, have observed a number of very significant irregularities in the University's proceedings, including significant violations of ORI/PHS regulations and guidelines. With the permission of Dr. Kay and her attorney (Robert Wrede, Esq.; Parker, Milliken, Clark, O'Hara & Samuelian, Los Angeles, CA), I am bringing these issues to your attention so you may consider whether a compliance review or other action by ORI is called for.

The most prominent procedural concerns relating to the University's investigation are the following:

(a) adventitious application of varying definitions of research misconduct, including the definition discarded by the PHS in 1989;

(b) questionable certification of regulatory compliance;

(c) failure to secure vital evidence;

(d) failure to provide adequate notice to Dr. Kay concerning the charges against her and the associated evidence (including failure to prepare even minimally adequate inquiry/investigation reports);

(e) failure to provide Dr. Kay reasonable opportunity to respond to the charges against her; and

(f) inequitable hearing procedures, including denial of full representation by counsel.

Background information on these issues follows.

The University pursued a highly suspect "misconduct" investigation of Dr. Kay based on opportunistic application of the "other serious practices" element of the existing misconduct definition as well as the very broad, no-longer-valid definition of misconduct long since discarded by the PHS. On a related matter, as recently as March, 1998, the University certified to ORI that its Policies and Procedures comply with the PHS regulation, even though the Policies and Procedures are centered on the now-invalid definition.

The University's inquiry into Dr. Kay's conduct (initiated in early 1997, following complaints by two junior laboratory staff, one of whom had recently resigned, about alleged laboratory mismanagement and fiscal improprieties) involved a wide variety of issues, including (a) alleged human subjects violations, (b) alleged violations of safe handling practices for radioactive and biohazardous research materials, (c) alleged failures of supervision, and (d) alleged fabrication of data. Concerning the two fabrication charges, the inquiry committee reported on April 15, 1997 that,

"The inquiry team was unable to find/obtain evidence required to support this allegation" and "The inquiry team was presented with no data to support this allegation" (Attachment #1).

On April 23, 1997, Michael A. Cusanovich, the University's Vice President for Research and Graduate Studies, wrote to Dr. Kay that,

"I am pleased to report that allegations of research fraud or data fabrication against you have not been substantiated" (Attachment 2).

Similarly, the same day, Dr. Cusanovich wrote to the Chairman of the University's Committee on Academic Freedom and Tenure (CAFT) that the inquiry committee "... did not support the allegations of fabrication of data by Dr. Kay ..." (Attachment 3).

Notwithstanding the inquiry committee's exoneration of Dr. Kay in the alleged fabrication of data, Dr. Cusanovich ordered that Dr. Kay

"... cease all activities involving the use of human subjects, radioisotopes, and biohazards" (Attachment 2)

and he called on CAFT to pursue a formal investigation of Dr. Kay, basing this call on the inquiry committee's "serious concerns about her research and laboratory safety procedures" (Dr. Cusanovich‚s words). Dr. Cusanovich's letter to CAFT cited the existing definition of "Misconduct in Science," making clear through his added emphasis, that he was relying on the "other practices that seriously deviate ...." element (emphasis supplied by Dr. Cusanovich) as a rationale for a misconduct investigation of Dr. Kay.

(I need hardly remind you of the longstanding concerns within the scientific community about the possible mischievous use of the "other serious practices" element to persecute a scientist who for one reason or another may have fallen out of institutional favor. In my view, this is precisely what has occurred in Dr. Kay's case, the most egregious and clear-cut instance I have seen of this kind.)

Further, also on the same day, Dr. Cusanovich wrote to Dr. Dorothy Macfarlane that the University's inquiry committee had

"... determined that probable cause exists to believe that Marguerite Kay ... has engaged in research misconduct" (Attachment 4).

Dr. Cusanovich's letter to Dr. Macfarlane mentioned only the laboratory safety and human subjects issues (issues that, as you know, are within the responsibility of other government agencies, not ORI). The letter made no mention of any unresolved allegations of fabrication or falsification of data. Dr. Macfarlane's response to the University indicated (a) it was ORI's understanding that the inquiry committee had dismissed the allegations of data fabrication and (b) ORI did not consider that its authority extended to the human subjects, biohazards, and other issues in the Kay case. Specifically, Dr. Macfarlane stated

"... ORI generally does not consider that violations of regulations enforced by other federal offices constitute scientific misconduct under the PHS definition" (Attachment 5).

As additional evidence that the University pursued Dr. Kay via adventitious application of varying definitions of research misconduct, as recently as December 1997, in its "Call for a Formal Hearing on the Allegations" relating to Dr. Kay, the CAFT laid out in its preamble this definition of "misconduct":

"... (1) fabrication, falsification, plagiarism, or other serious deviations from accepted practice in proposing, conducting or disseminating the results of research, scholarship or creative endeavor; (2) material failure to comply with federal and other requirements for protecting researchers, human subjects and the public or for ensuring the welfare of laboratory animals; and (3) failure to meet other material legal requirements governing research, scholarship and creative endeavors" (Attachment 6).

As you well-recognize, this definition incorporates the very broad, no-longer-valid definition under which the PHS operated until 1989. Importantly, this definition also fails to include the vitally-important limits on misconduct contained in the existing PHS definition, i.e.,

"It does not include honest error or honest differences in interpretations or judgments of data" (42 CFR Part 50.102).

It is my view that a number of the allegations regarding the work of Dr. Kay -- if they were shown to be factually correct -- would fall well within the parameters of this important condition.

Particularly significant in terms of institutional compliance, in its March 1998 Annual Report on Possible Research Misconduct (Attachment 7), the University affirmed to ORI that it has established a policy for responding to alleged research misconduct "... that complies with the PHS regulation ...." However, the 1995 policy cited in the University's Annual Report -- the policy applied by the University in its investigation of Dr. Kay -- does not comply with the PHS regulation, because, as noted above, the policy incorporates the no-longer-valid misconduct definition.

The University's reliance on varying misconduct definitions in its pursuit of Dr. Kay and its questionable certification of compliance are very troubling, particularly when viewed together with other significant defects, as cited below.

Contrary to ORI guidelines (ORI Model Policy for Responding to Allegations of Scientific Misconduct; ORI Model Procedures for Responding to Allegations of Scientific Misconduct) and simple common sense, the University failed to immediately secure original research data, subsequently invoking the missing data to the detriment of Dr. Kay.

There is no record that the University took any steps to secure the original research data upon initiation of its inquiry into Dr. Kay's research. Indeed, not until months later, in June of 1997, well after the April, 1997 start of the formal investigation, did the University obtain ORI's "List of Physical Evidence to be Secured."

In contrast, as demonstrated by an extensive documentary record, upon initiation of the University's inquiry, Dr. Kay promptly secured the data and records in her immediate possession. Also at this time, with increased urgency, Dr. Kay stepped up her ongoing efforts to recover other potentially relevant data, specifically, notebooks of the technician who had resigned the previous month and who was one of the two original informants against Dr. Kay. Despite Dr. Kay's repeated requests and contrary to the technician's obligation, she did not return her notebooks when she resigned. Dr. Kay sought the assistance of the Veterans' Administration facility where the technician carried out her laboratory work; Dr. Kay even requested that her attorney write University officials to urge them to recover the technician's notebooks. Despite these urgings, the University took no action for another month, and upon being rebuffed by the former technician (the University's own informant, who had been given "Whistleblower" status) the University apparently left off its efforts to obtain the missing data.

The significance of the missing data for Dr. Kay cannot be overstated:

(1) Dr. Macfarlane's April 25, 1997 letter to the University, responding to the University's "probable cause" notice to ORI, indicates that for ORI, the fact that data were missing was an important element in ORI's provisional acceptance of the case, i.e., according to Dr. Macfarlane, the fact that data were missing "... would usually support a decision that additional investigation is needed" (Attachment 5). The University, subsequently, cited the missing data as a "high priority" in its formal investigation.

(2) The data are important, possibly critical to Dr. Kay's defense, as even the CAFT chairman himself recognized, when he described these data as "... the crux of the first two charges" (Attachment 8).

(3) Given the University's own failings to secure the research data, it is especially troubling to see the University's two-edged, adventitious invocation of the missing data to the detriment of Dr. Kay. Note in particular the following excerpt from the CAFT "Notice of Hearing":

"In the most difficult instances, laboratory personnel's reports of lab results are directly contradicted by Dr. Kay's testimony of published studies. In most cases, with one significant exception, neither Dr. Kay, nor the lab personnel have been able to satisfactorily support either position by direct reference to the primary data in the laboratory notebooks or other records. Wherever possible the committee has tried to validate these statements with hard data or the considered opinion of experienced and trained professional researchers. Nonetheless, in some important instances the committee can only assume that the lack of supporting data either authenticates or disproves the allegation as appropriate. Where necessary, the committee has drawn reasonable inferences from available sources" (emphasis added; Attachment 6).

Whatever else one makes of this remarkably obscure passage, it is clear that in at least some instances, the absence of relevant data was construed as "authentication" of certain of the allegations against Dr. Kay.

The University has failed throughout to provide adequate notice to Dr. Kay concerning the charges and evidence against her. In particular, the University's "investigative" procedures have been cloaked in ambiguity, while its putative reports have failed entirely to meet ORI guidelines, much less basic requirements of due process. At the same, the University has imposed unreasonable conditions that have further hampered Dr. Kay's ability to respond.

The University's proceedings against Dr. Kay have moved along a steadily-shifting array of issues, from the originally-alleged financial improprities to the aforementioned alleged human subjects and biosafety violations to the alleged research fraud. Not-uncommon management issues amenable to straightfoward remedy (e.g., failure to document radiation safety training) have been denounced by University officials in hyperbolic verbiage that leaves one puzzled about their agenda, e.g.,

"... practices [that] threaten the credibility of research and science in the eyes of the community" (Attachment 9).

Yet, at the end of what the University calls its "investigatory phase of its process," the putative "specific charges of misconduct" lodged against Dr. Kay included the following seemingly minor issues, certainly falling well outside any reasonable definition of scientific misconduct:

"The publications are sloppily written ... Many of the publications are an ambiguous mixture of review and primary report ... Dr. Kay mismanaged her UA laboratory and employees in a manner so egregious and so outside the norms of University practice as to constitute serious misconduct ... Dr. Kay withheld pertinent information on employee tasks which makes it impossible to determine whether untrained employees worked with biohazardous materials ... She was seldom in the laboratory ... Dr. Kay promised pay at one level and then authorized a pay rate lower than promised ..." (Attachment 6).

To be sure, the charges also included a number of potentially very serious allegations of fabrication/falsification. But charge after charge was demonstrably ill-founded, as for example, charges that ascribe scientific claims to Dr. Kay when in fact her papers made no such claims. Other charges apparently arose out of misperceptions/misunderstandings. For example, in one instance, Dr. Kay was charged with selection of data, based on a paper's report that blood samples were drawn from ten subjects, while some of the paper's bar graphs are based on fewer subjects. What CAFT apparently missed in this regard is that several of the supposedly questionable graphs were based on brain, not blood samples!

The putative support for the University's fabrication/falsification charges is contained in a document that fails entirely to meet ORI's standards, as well as common-sense standards of reason and due process. ORI's model procedures, for example, require that an investigative report must, among other elements, include a list of the persons interviewed and "... the evidence secured and reviewed and the measures taken to ensure its security" (p. 32). In addition, for each allegation, the report must include background and analysis sections, the latter including

"all relevant statements, claims ... rebuttals, documents, and other evidence ... the source of each statement, claim, or other evidence should be cited" (p. 32).

Even a brief glance at the attached CAFT report shows that these evidence elements were almost never provided, and as a consequence, Dr. Kay was severely hampered in her efforts to prepare for the University's April 1998 hearing, the outcome of which will likely affect her entire research career. Consider, as just one example, the assertion in the CAFT report concerning one of Dr. Kay's papers, that

"... none of the probability values shown in figure legends or in footnotes to the tables are an accurate representation of the true probability values ... "

Incredibly, the report provided no substantiation for this sweeping assertion.

Where purported evidence was provided in the CAFT report, it most often comprised nothing beyond the unsubstantiated claims of an informant whose account apparently was uncritically accepted by the University.

Moreover, without substantiation, the report invariably put the most damaging construction (to Dr. Kay) on each piece of purported evidence. Thus, to cite one particularly egregious example, because one of Dr. Kay's papers provided a probability value for an experiment without detailing the underlying data, the report concluded, with no explanation, that the experiments in question "were not performed."

When the scanty, defect-ridden CAFT report was provided to Dr. Kay, at the end of 1997, it was provided as a final document. In other words, contrary to 42 C.F.R. Part 50 and ORI's Model Policy and Model Procedures, Dr. Kay was given no opportunity to review and respond to the report. Continuing this policy, the University did not identify its witnesses and proposed testimony for the April 1998 hearing until a mere 20 days prior to the hearing, and even then, only in the most superficial form (see Attachment 10). Further, during the hearing itself, the Chair announced that following the hearing, the panel would "... prepare a confidential report to the president of the University ... the confidential report is only shared with the president ... the confidential report is not shared with the respondent ..." (emphasis added). Thus, at this critical juncture, as throughout the entire process, Dr. Kay was denied reasonable opportunity to respond to specific, adequately substantiated charges, the associated evidence, and the conclusions based thereon.

The April 1998 hearing was notable for its procedural inequities, which without exception, worked to the disadvantage of Dr. Kay. In addition, and particularly troubling in light of these inequities, the University denied Dr. Kay full representation by counsel in the hearing.

The University's hearing was marked by inequitable, highly selective enforcement of quasi-legal purported rules, always to the detriment of Dr. Kay. The University set an arbitrary limit of five days for the hearing; four days were allocated to the University's case-in-chief; only one day was allocated to Dr. Kay's case. The hearing was purported to be an informal, information-gathering proceeding, the ultimate object of which was to permit the hearing panel to make its own, de novo judgement on the charges. But what actually occurred, e.g., the disallowance on exceedingly flimsy grounds of sworn, videotaped testimony supportive of Dr. Kay's work, from independent, international experts in relevant disciplines, was blatantly at odds with the professed intent of the hearing. In addition, the hearing panel's disallowance of testimony from Dr. Kay's experts, in contrast to the panel's acceptance of heresay, unsworn testimony and numerous unsworn documents from University witnesses, was clearly inequitable.

In light of the procedural inequities, a final and particularly troubling concern was the University's unyielding denial of Dr. Kay's right to full representation by counsel, a denial all the more striking in view of the legalistic trappings of the hearing process, e.g., cross-examination of witnesses. The denial of full representation is directly at odds with the hearing guidelines of the Scientific Integrity Panel of the HHS Departmental Appeals Board, which as you know, permit the full involvement (e.g., opening statements, cross-examination of witnesses, closing arguments, etc.) of attorneys for both the government and the accused. The biomedical research community fought hard for this due process protection at the federal level. It is astonishing and troubling that years later, an academic research institution adamantly denied this basic right to one of its own faculty members in so significant a proceeding.

(Dr. Kay's attorney has forcefully argued many of these due process issues. A recent representative letter to the University appears as Attachment 11.)

During our years at ADAMHA, you and I played significant roles in the formulation of the existing PHS policies and procedures for dealing with alleged research misconduct. In addition, as ADAMHA Administrator and NIMH Director, I reviewed the records and made final agency decisions in numerous cases of alleged misconduct. I am much concerned with the preservation of scientific integrity and the preservation of creativity and productivity in biomedical research. For these reasons, I am greatly troubled with the University of Arizona's actions in this matter, and believe they warrant the attention of ORI.

I have substantial additional information, including additional documentation, on the issues raised in this letter, which I would be happy to provide upon request. If you have questions or want to discuss these matters, I hope you will give me a call.

All best wishes,

Frederick K. Goodwin, M.D.
Director, Psychopharmacology Research Center, George Washington University
Director, Program on Medicine Science and Society, Ethics and Public Policy Center

 

Information on this page updated 10/31/99

Home ] Summary from 10/31/99 ] Court Info ] Letters of protest ] Letters Containing Info About This Case ]

1