Department of Health and Human Services (H.H.S.)
Departmental Appeals
Board
Appellate Division
DECISION AND FINDINGS OF FACT FOR THE PROPOSED DEBARMENTS OF ROBERT EDWARD
MCCAA, Ph.D.
Decision and Findings of Fact For Board
Docket No. 85-178
Decision No. 823
January 12, 1987
This is a decision on the proposal by the Deputy Assistant Secretary
for
Procurement, Assistance and Logistics (Debarring Official) of the
Department
of Health and Human Services (Agency, DHHS) to debar Robert
Edward McCaa,
Ph.D. (Respondent) from eligibility for, or involvement in,
grants and other
financial assistance awards (grants) from DHHS for a period
of three years.
This decision also contains findings of fact related to the
proposed debarment
of the Respondent from eligibility for contracts or
subcontracts (contracts)
with any federal department or agency. Under 45 CFR
76.15, this decision of
the Hearing Officer will be final with respect to
grants, unless, within 60
days, the Debarring Official reviews it. Under the
Federal Acquisition
Regulation, 48 CFR 9.406-3(d)(2), and 48 CFR
309.406-3(b), the findings of
fact will be considered by the Debarring
Official, who will make the final
decision with respect to
contracts.
The proposals to debar the Respondent from eligibility for
grants and
contracts stated that the Respondent had fabricated data,
published false
data, and used that false data in an application for a
renewal grant to the
National Heart, Lung, and Blood Institute (NHLBI). The
alleged scientific
misconduct arose in connection with research supported
directly or indirectly
under two grant awards from the NHLBI.
In
this decision, I present findings of fact with respect to the underlying
disputed issues. These include findings that the Respondent knowingly
reported
data and experimental procedures inaccurately and knowingly
reported research
results without any experimental foundation. I further
find that the
Respondent performed these acts over a period of several years
and that the
Respondent's judgment and general attitude toward scientific
research have not
changed since that time. These findings of fact relate to
the proposed
debarment for both grants and contracts.
For
debarment from grants, I also make conclusions of law, determining that
DHHS
has presented clear and convincing evidence of the existence of the
alleged
causes for debarment of the Respondent. I also conclude that the
evidence
clearly and convincingly shows that the Respondent is not presently
responsible to be a recipient of federal assistance.
Below, I
first state the decision on the proposed debarment. In the sections
that
follow, I present some regulatory provisions applicable to the proposed
debarments, a general procedural history of this dispute, a discussion of
the
administrative record and the weight and significance accorded some of
the
evidence, the findings of fact, a discussion of the analysis resulting
in
those findings, the conclusions of law applicable to the proposed
debarment
from grants, and a discussion of the analysis resulting in those
conclusions.
I will not directly address the proposed debarment from
contract awards, nor
will I present any conclusions of law with respect to
that matter, since that
would be outside of the scope of my authority as the
Hearing Officer for this
matter. 45 CFR 76.14, 76.15; 48 CFR 9.406-3(d),
309.406-3(b).
I. Statement of Debarment Decision
The Respondent in this case was a professor of physiology and
biophysics at
the University of Mississippi Medical Center (UMMC). He
received federal grant
funds from the NHLBI, a part of the National
Institutes of Health (NIH), as a
researcher under grants awarded prior to
1977, and as the principal
investigator from 1977 until May 10, 1985 under
grant award RO1-HL- 09921,
"Mechanisms Related to Congestive Heart Failure."
The Respondent's research
specialty was in the area of hypertension and
blood chemistry.
In 1982, NIH and UMMC received anonymous allegations
that data in three
published articles were "probably totally fabricated."
The Respondent was the
only common co-author on these papers. These
allegations, and other
allegations connected with the Respondent's
publications, were investigated
both by UMMC and by a panel of prominent
scientists appointed by NHLBI. On the
basis of the report of the NHLBI
panel, an NIH senior staff committee
recommended that the Respondent should
be debarred from further grants and
contracts. The Debarring Official issued
the proposals to debar the
Respondent, and I was appointed as the Hearing
Officer.
This is the first contested debarment of a research scientist
at DHHS. While
I used the general requirements for debarment proceedings in
other contexts as
a guide, I gave special consideration to ensuring the
fairness of this
proceeding because the Respondent is an individual who has
been dependent upon
federal research grant funds for his livelihood. I held
a full, evidentiary
hearing, and I provided the Respondent with numerous
opportunities to submit
written evidence and argument. I then made
determinations based upon the full
record before me.
I placed the
burden of proof on DHHS to establish the causes for debarment
with evidence
sufficient to meet the regulatory standards at 45 CFR Part 76,
48 CFR 9.406
and 48 CFR 309.406. Since no detailed statutory or regulatory
codes
governing the conduct of scientific research were submitted to me, I
based
my decision only on causes for which DHHS established the violation of a
regulatory condition, of a standard of conduct affecting responsibility, or
of
a term or condition of the Respondent's grant.
My decision that
the Respondent should be debarred is based primarily upon
my conclusions
that the Respondent violated two fundamental standards of
conduct for
research scientists receiving federal grant funds. The first
standard is the
general duty of scientists to accurately report the
procedures, results, and
other observations from experiments performed. I was
convinced by the
testimony and other submissions from prominent scientists of
the paramount
importance of accuracy and honesty to the experimental method of
scientific
advancement. The second standard is the duty of the principal
investigator
on a federal research grant to bear responsibility for the
scientific and
technical conduct of the research program under the grant. This
duty is
implicit in the concept of a principal investigator, and is expressed
in the
regulations at 42 CFR Part 52, in the Public Health Service (PHS)
Grants
Policy Statement applicable to NIH grants, and is also incorporated in
the
grant by the statement which a principal investigator signs on the grant
application agreeing to accept this responsibility.
The Respondent
violated these standards of conduct with his careless
disregard for accuracy
in scientific reporting and his lack of respect for the
experimental method.
The record contains clear and convincing evidence that
the Respondent
fabricated and falsified data, published that data in
scientific journals as
the results of experiments actually performed, and used
that false data to
support applications for grant funds from the NHLBI. The
Respondent's
pattern of "scientific misconduct" and lack of knowledge about
generally
accepted research practices were reflections of an overall lack of
integrity
and honesty. [FN1]
I based my conclusions on the pattern of conduct
which became evident as the
record developed, not on any single element. I
found most persuasive the
evidence that, in one paper, the Respondent
reported data for the levels of
certain blood chemicals for a period during
which no blood samples were taken
from the experimental animals. Other
evidence indicated that the Respondent
reported as experimental procedures
or results inaccurate information based
upon his expectations or
suppositions, rather than the actual occurrences,
when he did not know what
the actual procedures or results were from an
experiment. Whether or not his
expectations or his suppositions were accurate,
the record indicated that
reporting information in that way is contrary to
generally accepted
practices for the reporting of scientific research, unless
the reporting is
identified as hypothetical or otherwise unsupported by
experimentation.
The record before me indicates that one reason
why the Respondent acted in
this manner is that he reported on experiments
in which he had limited
involvement and knowledge, and did not consult with
those who had actually
conducted the experiments to ensure the accuracy of
his reports.
The pattern which emerged from the record was one of
intentional deception
of both the public and the federal government in order
to enhance the
Respondent's reputation and his likelihood of receiving grant
funds. The
Respondent went beyond mere self-promotion in reporting on
experimental
procedures which had not been performed and results which had
not been
obtained. Furthermore, in some instances, he reported exaggerated
numbers of
subjects which would mislead readers into believing that the
results obtained
were more significant and accurate than they were in
fact.
In determining that the Respondent should be debarred, I have
considered the
strong public interest in protecting the integrity of the
scientific research
process, in preventing the potential abuse of federal
funds, and in minimizing
the potential for continued fraud on the grant
award process. The record
indicates that the Respondent's actions took place
over a period of years, and
I have not found any reason to believe that the
pattern could not continue. I
have not found that the Respondent's actions
had any reasonable explanation
due to unusual circumstances. I have not
found any evidence of remorse or
changed attitude sufficient to change my
conclusion that the Respondent is not
presently responsible as a recipient
of federal funds. In fact, during the
course of this proceeding, I have been
persuaded of the opposite by the
Respondent's relentless defense of actions
in contravention of generally
accepted standards of conduct.
Thus,
for the reasons discussed more fully below, I have determined that the
interests of the public and of the federal government require that the
Respondent, Robert Edward McCaa, Ph.D. be debarred for the full three-year
term proposed by DHHS, reduced, as agreed by DHHS, by approximately six
months
in recognition of certain unforeseen delays in this particular
proceeding.
Accordingly, the debarment term runs from July 15, 1986.
II. Regulatory Authority and Standards
This proceeding is held under the authority of regulations governing
administrative debarments from grants and contracts. See, generally, 45 CFR
Part 76; 48 CFR Parts 9.4 and 309. By means of an administrative debarment,
institutions or individuals are excluded from eligibility for grants or
contracts with the federal government for a specified period of time.
Administrative debarments are to protect the interests of the government and
are not punitive. 45 CFR 76.1; Gonzalez v. Freeman, 334 F.2d 570
(D.C.Cir.1964); see, generally, S. Horowitz, Looking for Mr. Good Bar: In
Search of Standards for Federal Debarment, 14 Pub.Cont.L.J. 58
(1983).
Under 45 CFR 76.10, an institution or individual may be
debarred from
eligibility for grants based on clear and convincing evidence
showing any of
the several enumerated causes cited in the proposed debarment
at issue here:
(d) Serious violation of the applicable
statute, regulations, or other
terms and conditions of a previous award of
financial assistance;
(e) A record of serious
unsatisfactory performance (or failure to perform)
under one or more prior
awards of financial assistance, contracts or
subcontracts, as determined
under the terms and conditions or specifications
of the prior awards, except
that unsatisfactory performance (or failure to
peform) caused by acts beyond
the control of the institution or individual
shall not be considered a basis
for debarment;
(g) Any other cause significantly affecting
responsibility as a recipient
or participant under a Federal program of
sufficiently serious nature as
determined by the Secretary to warrant
debarment.
Debarment is not mandatory upon a determination that the
causes for
debarment exist. A determination to debar is discretionary and
"shall be
rendered solely in the best interests of the Government" after
consideration
of the entire administrative record, as well as the "degree of
seriousness of
the offense, violation, failure, or inadequacy of
performance," and the
existence of any other mitigating factors. 45 CFR
76.11.
Under 48 CFR 9.406-2 the Debarring Official may debar a
contractor from
eligibility for contracts based upon a preponderance of the
evidence showing
any of several enumerated causes, including the cause cited
in the proposed
debarment at issue here:
(c) Any other
cause of so serious or compelling a nature that it affects
the present
responsibility of a Government contractor or subcontractor.
Such a
determination is discretionary, based on "the public interest," and
should
be made after consideration of "the seriousness of the contractor's
acts or
omissions and any mitigating factors." 48 CFR 9.406-1.
The regulations
for debarment from grants provide for an opportunity for a
hearing upon the
Respondent's request. 45 CFR 76.14(b). The regulations for
debarment from
contracts provide in general for an opportunity to be heard and
specifically
for some type of fact-finding proceeding where there is a
"genuine dispute
over [material] facts." 48 CFR 9.406-3(b)(2), 309.406-3(b).
[FN2]
III. Procedural History
In December 1982, the NIH and NHLBI established a panel of "senior
investigators with experience in hypertension" (NHLBI panel) to investigate
allegations that the Respondent fabricated data in research papers supported
by NIH grant funds. Ex. 60, Tab 1, p. 1 [FN3] The allegations first surfaced
in anonymous letters sent in March 1982 to the Acting Director of the NHLBI,
the Dean of the University of Mississippi Medical Center (UMMC), and the
Chairman of the Department of Physiology and Biophysics at UMMC. Ex. 60, Tab
1, p. 1. In September 1982 the UMMC had issued a report by senior faculty
members which examined the allegations made in the anonymous letters and
contained additional allegations of improper scientific conduct by the
Respondent. Ex. 60, Tab 1, pp. 6-7; Ex. 196.
The NHLBI panel
issued its report (the NHLBI report) on July 20, 1984. The
panel concluded
that the Respondent had engaged in "serious scientific
misconduct" and
recommended that the Respondent be debarred from eligibility
for financial
assistance and other awards and that the Respondent's current
grant be
terminated. Ex. 60, Tab 1, p. 26.
A committee of senior NIH staff
reviewed the NHLBI panel report as well as
additional comments submitted by
the Respondent and University officials. On
October 17, 1984, two NIH
officials met with the Respondent, at his request.
The NIH senior staff
committee findings were presented in an April 30, 1985
memorandum to the
Director of NIH. Ex. 60, Tab 2. This committee concurred
with the NHLBI
panel and, inter alia, recommended debarment proceedings.
The Director
of NIH, on May 3, 1985, notified the Respondent that he had
recommended to
the Secretary of DHHS that DHHS debar the Respondent from
eligibility for
federal grants and contracts. Ex. 60, Tab. 3. On August 15,
1985, the
Debarring Official notified the Respondent of the proposed
debarments at
issue, in separate letters for debarment from federal grants and
for
debarment from federal contracts. [FN4] Exs. 67, 68.
The proposed
debarment from grants would bar the Respondent from receiving
discretionary
grants from DHHS, and prohibit the Respondent from serving or
participating
in the conduct of any grant, for a period of three years. The
debarment
would be publicly announced in the Federal Register. 45 CFR 76.17.
The
proposed debarment from contracts would bar any federal agency from
accepting or soliciting bids and awarding or renewing contracts with the
Respondent, absent a compelling reason, for a period of three years. The
Respondent's name would be placed on the Consolidated List of Debarred,
Suspended, and Ineligible Contractors, maintained by the General Services
Administration.
The proposed debarments were based on "serious
scientific misconduct,"
including fabrication of data, publication of false
data, and use of that
false data in grant applications to the NHLBI. The
alleged misconduct, the
Debarring Official stated, reflected a lack of
integrity and honesty, as well
as a lack of knowledge about generally
accepted scientific research practices.
These failings, the Debarring
Official asserted, indicated the absence of
present responsibility as a
recipient of grants and contracts.
The Respondent timely submitted to
the Debarring Official written
information and argument in opposition to the
proposed debarments and
requested a hearing. The Debarring Official
requested that the Departmental
Grant Appeals Board hold the hearing and
issue both a decision on whether the
Respondent should be debarred from
eligibility for discretionary financial
assistance and findings of fact on
all disputed issues related to debarment
from eligibility for government
contracts. See Memorandum from Debarring
Official to Chair, Departmental
Grant Appeals Board, dated October 9, 1985.
The Chair of the
Departmental Grant Appeals Board designated me as the
Hearing Officer. Both
parties had the opportunity to contribute documents to
the record, and I
presided over a hearing during which both parties presented
witnesses and
had the opportunity to cross-examine opposing witnesses. There
have been
opportunities to present briefs on legal and factual isses both
before and
after the hearing.
IV. The Decision Process
This decision is based upon the entire documentary record as well as
the
hearing testimony presented by the parties. This includes information
developed in the NHLBI panel inquiry and the UMMC investigations described
above. Using these sources along with the submissions made by each party, I
have made independent findings with respect to all contested
facts.
I placed the burden of proof on the Agency to present evidence
meeting the
regulatory standards in support of the proposed debarment.
Although I assigned
no presumptive weight to the conclusions of the prior
investigatory bodies, I
found the NHLBI panel report to represent a reliable
expert opinion on the
generally accepted standards of scientific conduct and
on the significance of
the alleged actions to the scientific community.
Also, while I did not regard
the conclusions stated in the NHLBI report
alone to establish the existence of
the causes for debarment, their
conclusions regarding the falsified and
fraudulent data in the Respondent's
work are highly significant. These
conclusions corroborate those I reach in
the findings of fact after analyzing
the record as a whole developed during
this proceeding. I did not accord the
same weight to the UMMC investigatory
findings as to the NHLBI report,
primarily because of the greater interest
in the outcome by the UMMC faculty.
Also, the greater fact-finding
capability available in the NHLBI process and
in this proceeding made it
unnecessary to rely on the UMMC report.
Although the Respondent
alleged that the NHLBI panel merely rubber- stamped
the UMMC investigatory
findings, this allegation is unfounded. The NHLBI panel
undertook an
extensive independent investigation; the NHLBI panel reviewed the
entire
documentary record, conducted interviews and site visits, solicited
written
statements, and distributed a draft report for comments from
interested
parties. Ex. 60, Tab 1, pp. 4-5. The panel was highly
distinguished, as
discussed above, and did not limit itself to the UMMC
investigatory
findings. Id., p. 3. The NHLBI panel report appeared to be a
conservative
document, rejecting some findings because of a lack of
substantial evidence,
and adopting others which were more clearly established
by the evidence.
See, e.g., Ex. 60, Tab 1, pp. 11-12.
In weighing the evidence before
me, it was necessary for me to assess the
credibility of the witnesses
present at the hearing. My assessments are stated
below, when relevant to my
evaluation of the evidence presented. In general, I
found the Respondent to
be less credible than some of the other witnesses
because of inconsistencies
between his oral testimony and his prior written
statements submitted to
UMMC and NIH, and because of his generally imprecise
and rambling testimony.
The Respondent overgeneralized scientific issues which
other witnesses
presented in greater detail, tended to exaggerate, and would
sometimes avoid
answering questions directly unless pushed to do so by the
questioner.
Furthermore, the Respondent recounted his interactions with others
at UMMC
with details, including quoted language, which I did not find credible
considering that the events in question occurred, for the most part, eight
or
nine years ago. [FN5]
The Respondent's efforts to shift the
blame to others with regard to the
alleged incidents of scientific
misconduct as well as the Respondent's efforts
to discredit others on the
faculty by making allegations of fabrication and
falsification of data in
their work diminished the persuasiveness of
Respondent's evidence in
general. [FN6] In examining the articles which the
Respondent alleged to
contain fabricated information, I became aware of
differences in the
presentation of information in these articles from the
presentation in the
Respondent's articles. This examination actually
highlighted the flaws in
the use and presentation of data present in the
Respondent's work.
The testimony of Drs. Bengis, Coleman, and Hall was in marked contrast to
the Respondent's testimony. All three impressed the Hearing Officer with
their
scientific expertise and with the care and precision with which they
testified
about the scientific aspects of this case. These witnesses were
forthright,
especially with regard to any limits of their recollection,
which further
enhanced the credibility of their testimony in general.
V. Findings of Fact
Unless otherwise specified, the Agency proved the following facts both
by
clear and convincing evidence and by a preponderance of the evidence.
A. GENERAL BACKGROUND
1. THE RESPONDENT, DR. ROBERT EDWARD MCCAA, RECEIVED FUNDING FROM THE
NATIONAL HEART, LUNG, AND BLOOD INSTITUTE, PART OF THE NATIONAL INSTITUTES
OF
HEALTH, WHILE EMPLOYED AT THE UNIVERSITY OF MISSISSIPPI MEDICAL CENTER
FOR
RESEARCH ACTIVITIES UNDER THE FOLLOWING GRANT AWARDS FOR THE PERIODS
SPECIFIED:
a. GRANT AWARD P01-HL-11678, ENTITLED
"CARDIOVASCULAR DYNAMICS AND THEIR
CONTROL," FOR WHICH DR. ARTHUR C. GUYTON,
THE CHAIRMAN OF THE UMMC DEPARTMENT
OF PHYSIOLOGY, WAS PRINCIPAL
INVESTIGATOR, FROM MAY 1, 1974 TO JANUARY 31,
1978. FROM JANUARY 31, 1978 TO
JANUARY 31, 1984, THE RESPONDENT WAS LISTED AS
PROVIDING EFFORT, BUT
RECEIVED NO SALARY UNDER THE GRANT AWARD.
b. GRANT AWARD
R01-HL-09921, "MECHANISMS RELATED TO CONGESTIVE HEART
FAILURE," FROM WHICH
THE RESPONDENT RECEIVED SALARY SUPPORT FROM DECEMBER 1,
1971 UNTIL MAY 10,
1985. FROM DECEMBER 1, 1971 UNTIL DECEMBER 1, 1977, DR.
CONNIE S. MCCAA, THE
RESPONDENT'S WIFE, WAS THE PRINCIPAL INVESTIGATOR UNDER
THIS GRANT. THE
RESPONDENT WAS PRINCIPAL INVESTIGATOR FROM DECEMBER 1, 1977
UNTIL MAY 10,
1985.
THE RESEARCH AND PAPERS AT ISSUE IN THIS CASE WERE SUPPORTED BY FUNDS
FROM
GRANT AWARD R01-HL-09921.
2. THE RESPONDENT HAS NOT RECEIVED
FUNDING FOR RESEARCH ACTIVITIES SINCE MAY
10, 1985, WHEN THE DEAN OF UMMC
REQUESTED THAT THE RESPONDENT IMMEDIATELY
CEASE FURTHER WORK, AND
DISCONTINUE EXPENDITURE OF FUNDS, UNDER GRANT AWARD
R01-HL- 09921.
3. THE EVIDENCE PRESENTED IN SUPPORT OF THE PROPOSED DEBARMENT CONCERNS
PRIMARILY THE PERIOD OF 1976 THROUGH THE PRESENT. DURING THIS PERIOD, THE
RECORD CONTAINS GENERAL EVIDENCE THAT THE FOLLOWING MAY HAVE AFFECTED THE
RESPONDENT'S ACTIONS:
a. PRESSURES TO PUBLISH RESEARCH
RESULTS TO SUPPORT THE RESPONDENT'S
APPLICATIONS FOR GRANT FUNDS, ENSURE HIS
CONTINUED EMPLOYMENT AT UMMC,
MAINTAIN HIS ACCESS TO PROPRIETARY DRUGS, AND
ENHANCE HIS GENERAL STANDING IN
THE SCIENTIFIC
COMMUNITY.
b. PROBLEMS IN THE RESPONDENT'S PERSONAL
LIFE.
c. PROBLEMS RELATED TO CHANGES IN THE PHYSICAL
LOCATION OF THE
RESPONDENT'S LABORATORY FACILITIES.
Discussion of General Background
The facts in this section of my findings were not contested by the
parties.
I note that, during the period in which the Respondent received
funding from
NHLBI, he was an assistant professor of physiology in the UMMC
Department of
Physiology and Biophysics from July 1971-June 1974, an
associate professor
from July 1974-June 1977, and a full professor from July
1977 until his
employment was terminated on October 17, 1985. Ex.
74.
The record contains references, only generally explained, to
personal
problems of the Respondent during the time period when the series
of disputed
actions started. Dr. Guyton, in a written response to questions
from the NHLBI
panel, referred to "the death of his [the Respondent's]
father, sickness and
ultimate death of his mother to whom he was deeply
devoted, his marriage going
on the rocks for awhile, travelling far too much
to meetings and site visits,
paying far too little time to his research work
and yet having the necessity
to put together new research data all the time
to support the travelling and
reports, and difficulties he was having
maintaining his research grant with
NIH." Ex. 160, p. 23. Dr. Langford also
mentioned, in an interview with NHLBI
panel members, that the Respondent
during that time period "tried to do too
much" and had a variety of personal
problems. Ex. 158, p. 2.
Additionally, uncontested evidence was
presented that contributing factors
to some loss or confusion regarding
records of research data may have been the
relocation, in late September
1977, of the Respondent's laboratory facilities
within the UMMC facility and
the lack of sufficient office and file space. See
Ex. 155, p. 3; Tr. Jan.
17, 1986, pp. 73-76.
B. FACTS RELATED TO THE PAPER, "THE ROLE OF ALDOSTERONE IN
EXPERIMENTAL
HYPERTENSION."
AUTHORSHIP
4. IN FEBRUARY 1978, THE RESPONDENT SUBMITTED FOR PUBLICATION IN THE
JOURNAL
OF ENDOCRINOLOGY A PAPER ENTITLED "THE ROLE OF ALDOSTERONE IN
EXPERIMENTAL
HYPERTENSION." THIS PAPER WAS SUBSEQUENTLY PUBLISHED IN VOLUME
81 OF THE
JOURNAL OF ENDOCRINOLOGY AT PP. 69-78 WITH THE AUTHORS LISTED AS
R.E. MCCAA
(THE RESPONDENT), C.S. MCCAA, R.G. BENGIS AND A.C. GUYTON. EX.
15.
5. THE RESPONDENT WROTE THE "METHODS," "RESULTS" AND PORTIONS OF
THE
"DISCUSSION" SECTIONS OF THIS PAPER. HE WAS SOLELY RESPONSIBLE FOR
THESE
SECTIONS AND DID NOT CONSULT WITH ANY OF THE OTHER AUTHORS WHILE
WRITING THESE
SECTIONS.
Discussion of Authorship
The facts related to authorship of this paper (the MMBG Paper) were
not
contested by the parties. The paper was written to accompany a lecture
by the
Respondent at the Tait Symposium in London in November 1978. The
Respondent's
wife, Dr. Connie McCaa, wrote the introduction and parts of the
discussion
sections of the paper, but she did not need to refer to the data
to write
these sections. Tr. Apr. 2, 1986, pp. 127 and 133. See Ex. 84. I
will address,
in the following discussion, the findings of fact in points
6-13, all of which
were contested by the parties. The discussion will be
referenced to the
appropriate finding of fact, but will not repeat the
factual conclusion.
CIRCUMSTANCES OF AUTHORSHIP
6. AT THE TIME THE PAPER WAS SUBMITTED, THE RESPONDENT WAS UNDER
CONSIDERABLE PRESSURE TO COMPLETE THE PAPER.
Discussion of Circumstances of Authorship
Dr. Connie McCaa testified that the paper was not ready for submission
at
the time the Respondent arrived in London to present the paper at the
Tait
Symposium. Tr. Apr. 2, 1986, pp. 126-129. She added that the Respondent
only
completed the paper while in London with a typewriter borrowed from the
Royal
Society of Medicine. Id.
The record indicates that the paper
and the presentation at the Tait
Symposium were important to the advancement
of the Respondent's career. The
Respondent's subsequent citation of the
paper in his grant applications
indicates that the research was used as
support for the Respondent's future
requests for grant funds. See Ex. 227.
The Respondent's testimony about his
relationship with Squibb Pharmaceutical
Company indicates that, although there
was no formal agreement, he was
expected to conduct research studies when he
received drug samples. Tr. Jan.
17, 1986, pp. 21-23. Furthermore, the record
indicates that both the paper
and the presentation advanced the Respondent's
general standing in the
scientific community.
BASIS FOR THE MMBG PAPER
7. THE "METHODS" AND "RESULTS" SECTIONS PURPORT TO REPORT DATA DERIVED
FROM
EXPERIMENTS ACTUALLY PERFORMED ON LABORATORY RATS TO SHOW THE RESULTANT
CHANGES IN THEIR BLOOD CHEMISTRY. THE EXPERIMENTS INVOLVED A CONTROL GROUP
OF
SUBJECTS WHICH WERE KEPT UNDER NORMAL CONDITIONS, AND FIVE GROUPS OF
SUBJECTS
WHICH WERE SUBJECT TO SPECIAL DIETARY OR SURGICAL TREATMENT. EACH
GROUP
(INCLUDING THE CONTROL GROUP) WAS SELECTIVELY ADMINISTERED DOSES OF
SQ-14,225
(CAPTOPRIL), AN EXPERIMENTAL DRUG DEVELOPED BY THE SQUIBB
PHARMACEUTICAL
COMPANY, FOR A SEVEN-DAY EXPERIMENTAL PERIOD. EX.
15.
8. THE ONLY EXPERIMENTS WHICH THE RESPONDENT PURPORTED TO REPORT
IN THIS
PAPER WERE CONDUCTED BY DR. ROY G. BENGIS (THE BENGIS EXPERIMENTS),
WITH
ADVICE AND TECHNICAL ASSISTANCE FROM OTHERS, AS PART OF THE COURSE OF
STUDY
FOR HIS DOCTORATE. THESE EXPERIMENTS WERE REPORTED IN DR. BENGIS'
DOCTORAL
THESIS, TWO ARTICLES WRITTEN PRIMARILY BY DR. BENGIS AND DR. THOMAS
G.
COLEMAN, AND TWO ABSTRACTS WRITTEN BY THE RESPONDENT. THE RESPONDENT
INDEPENDENTLY CONDUCTED NO EXPERIMENTS INVOLVING RATS WHICH COULD HAVE
PRODUCED ANY DATA REPORTED IN THIS PAPER.
Discussion of the Basis for the MMBG Paper
7. The MMBG paper itself describes the methodology purportedly used to
obtain the results reported. Ex. 15, p. 70. The paper references the work of
Drs. Bengis and Coleman for a more detailed description of the experimental
design. The parties did not dispute the number of groups of subjects or the
use of captopril. Id. These descriptions are consistent with this finding.
See
Agency Hearing Ex. C; Respondent Hearing Ex. B.
8. The
Respondent admitted that the data contained in this paper resulted
only from
experiments conducted by Dr. Bengis. Ex. 27, p. 3. The Respondent
has never
asserted that he conducted any independent experimentation with rats
upon
which any data reported in this paper was based. This finding relates
only
to the stages of the experiment directly involving rats, not to the
subsequent performance or non-performance of tests on blood samples taken
from
the subjects in the Bengis experiments.
The Respondent did
not contest that the experiments, which he purported to
report in this
paper, were the same experiments reported in Dr. Bengis'
thesis: "The
Effects of Prolonged Blockade of the Renin-Angiotensin System on
Blood
Pressure Control in Rats," Agency Hearing Ex. C. These experiments were
also
the subject of two other articles: "Long-Term Blockade of Angiotensin
Formation in Various Normotensive and Hypertensive Rat Models Using
Converting
Enzyme Inhibitor (SQ 14,225)," Bengis, Coleman, Young, and McCaa,
Respondent
Hearing Ex. B; and "Antihypertensive Effect of Prolonged Blockade
of
Angiotensin Formation on Benign and Malignant, One-and Two-kidney
Goldblatt
Hypertensive Rats," Bengis and Coleman, Ex. 14. The Respondent
argued that
these other sources were not complete; that they were either
based on
preliminary data or did not report the entire set of procedures
conducted and
results obtained. Drs. Bengis and Coleman agreed that the
earlier article,
Respondent Hearing Ex. B, was based on preliminary data and
covered only some
of the data from the Bengis experiments. But both he and
Dr. Bengis firmly
stated that the other article, and Dr. Bengis' thesis,
accurately reported the
procedures and final data from the experiments
conducted by Dr. Bengis. Tr.
Jan. 15, 1986, pp. 6, 16, 32, 192.
As
I will discuss in more detail as I examine the contrasts between the
Respondent's reporting of the Bengis experiments and the reporting by Drs.
Bengis and Coleman, I find the reporting by Drs. Bengis and Coleman to be
more
credible and complete, at least with regard to the parts of the
experiment in
which Drs. Bengis and Coleman were directly involved, than the
Respondent's
reporting of the experiments.
Neither party to this
proceeding contested the fact that the Respondent
submitted for publication
two abstracts using preliminary data given to him by
Dr. Bengis: "Arterial
Pressure Response in Experimental Renovascular
Hypertensive Rats During
Long-Term Inhibition of Angiotension I Converting
Enzyme With SQ-14,225,"
Ex. 9, and "Arterial Blood Pressure Response to Long-
Term Inhibition of
Angiotensin I Converting Enzyme with SQ-14,225 in
Experimental Renovascular
Hypertensive Rats," Ex. 10.
At several points in the record, the
Respondent suggested the possibility
that data from blood samples from
experiments conducted by Bennye Henderson, a
graduate student working
primarily with hypophysectomized rats, may have been
confused and thought to
be data from samples from the Bengis experiments. See,
Ex. 73, p. 2; Ex.
155, pp. 2-4; Tr. Jan. 17, 1986 (cont.) pp. 70-78. Some of
Ms. Henderson's
work may have overlapped with Dr. Bengis' work; her first
experiments
included blood samples for rats kept under normal and sodium-
depleted
conditions and treated with captopril. The Respondent was not certain
that
the values for her later work with hypophysectomized rats would have been
the same as those reported for experimental groups in the Bengis
experiments,
although he stated that they might be within the same range.
Id., pp. 72-76.
Dr. Coleman testified that the values would not have been
the same. Tr. Jan.
16, 1986, p. 5. Furthermore, the Respondent stated that
the samples had been
clearly marked as Dr. Bengis' samples, and he did not
rely on any defense
based on a confusion of samples. Id., p. 71. In sum, the
testimony was that
confusion of the samples, while possible, would not have
resulted in all the
values reported in the MMBG paper, and the Respondent
did not attribute those
values to confusion of samples.
METHODOLOGY OF THE MMBG PAPER
9. THE SECTIONS WRITTEN BY THE RESPONDENT DID NOT ACCURATELY REPORT
THE
PROCEDURES USED BY DR. BENGIS TO CONDUCT THE
EXPERIMENTS.
a. THE SECTIONS WRITTEN BY THE RESPONDENT DID
NOT ACCURATELY REPORT EITHER
THE NUMBER OF SUBJECTS FOR EACH DATA POINT
REPORTED OR THE NUMBER OF SUBJECTS
IN EACH EXPERIMENTAL
GROUP.
b. THE SECTIONS WRITTEN BY THE RESPONDENT DID NOT
ACCURATELY REPORT THE
LENGTH OF THE CONTROL PERIOD USED FOR EACH
EXPERIMENTAL GROUP.
c. THE SECTIONS WRITTEN BY THE
RESPONDENT DID NOT ACCURATELY REPORT THAT
NO BLOOD SAMPLES WERE OBTAINED
FROM THE EXPERIMENTAL SUBJECTS DURING PERIODS
OF CAPTOPRIL
ADMINISTRATION.
Discussion of the Methodology of the MMBG Paper
In this paper, the
Respondent presented a general description of the
experimental procedures.
Ex. 15, p. 70. This description conflicts with the
descriptions of the
procedures contained in papers done by Drs. Bengis and
Coleman. Agency
Hearing Ex. C; Respondent Hearing Ex. B; Ex. 14. The
Respondent's
description contains a larger number of test subjects, a 2-week
control
period rather than a control period of only 3 or 4 days, and a
description
of blood samples obtained during the period of captopril
administration when
Drs. Bengis and Coleman state unequivocably that no blood
samples were
taken. In general, I found the descriptions by Drs. Bengis and
Coleman to be
more credible because of the direct involvement of Drs. Bengis
and Coleman
in conducting the non-laboratory phases of the experiments.
Furthermore, Dr.
Bengis' doctoral thesis was highly detailed and appears to
have been subject
to review by a panel of advisers. Agency Hearing Ex. C.
Although the
Respondent alleged that these witnesses were biased against him
because of
either individual motivations or a conspiracy, there was little
evidence
which supported this view. In particular, the Respondent provided no
adequate explanation of why Dr. Bengis was motivated to discredit him, or
would participate in any conspiracy. The Respondent implied that Dr. Bengis'
testimony was influenced, during the initial investigation stage, by
pressure
exerted by the Chairman of the UMMC Department of Physiology, Dr.
Arthur
Guyton, and the UMMC counsel. In April 1982, Dr. Guyton met with Dr.
Bengis.
Also in 1982, the UMMC counsel sent Dr. Bengis a copy of Section III
of the
draft UMMC report, containing Dr. Coleman's account of the facts
related to
the MMBG paper.
But the Respondent never developed any
evidence which would indicate that
these events affected Dr. Bengis'
testimony. Dr. Bengis was not a faculty
member at UMMC. He had obtained his
degree several years earlier, and was not,
apparently, beholden to UMMC or
any particular faculty members. He is not
practicing in the fields of
physiology or medical research, or even living in
the United States. There
is no evidence to suggest that he would be motivated
to lie in order to
protect his reputation, to preserve good relations with the
UMMC Department
of Physiology, to obtain grant funds from the United States,
or to avoid any
prosecution or other action by the United States.
Furthermore, Dr.
Bengis' statements throughout the record are consistent;
the Respondent's
allegation that Dr. Bengis' recollection changed after being
exposed to Dr.
Coleman's account of the rat experimentation in the draft UMMC
report is not
supported by the record. Respondent's Post-Hearing Brief, p. 72.
Dr. Bengis
was very frank about the limitations of his memory, but was clear
and
consistent about significant details. Moreover, Dr. Guyton's account of
his
meeting with Dr. Bengis in April 1982 does not indicate any change in Dr.
Bengis' recollection. Dr. Guyton recalled some gaps in Dr. Bengis' memory,
but
not in any areas in which I will rely on Dr. Bengis' testimony. See Ex.
160,
p. 11. Dr. Bengis, himself, admitted that his recollection is spotty of
events
of over seven years past.
In general, I find the
Respondent's explanations of the reasons he reported
procedures
inaccurately, when the record is clear concerning such
inaccuracies, not to
be credible. In light of the fact that, in his paper, the
Respondent cited
Dr. Bengis' thesis and the preliminary paper by Drs. Bengis
and Coleman, I
cannot give much weight to his statements that he should not be
held
responsible for discrepancies with these papers because he was unaware of
the content of these papers, including any changes in procedures, of precise
numbers of test subjects, or of details such as control periods. [FN7] All
of
these details are clearly stated in the work of Drs. Bengis and Coleman
which
the Respondent cited in the methods section of the MMBG paper. Ex. 15,
p. 70.
Although the Respondent alleged that he had sufficient independent
knowledge
of the Bengis experiments to write the MMBG paper, these citations
signify the
Respondent's admission that Drs. Bengis and Coleman had a more
direct
knowledge of the details of the experiments. The citations
acknowledge that
the work of Drs. Bengis and Coleman should be regarded as
the authoritative
account of the experiment and that the Respondent accepted
responsibility for
actual or constructive knowledge of the contents of the
cited works. Although
the Respondent knew of these sources and may have
regarded them as
authoritative, the Respondent reported the experimental
procedures based on
what he recalled or surmised, without bothering to check
either the published
sources or to ask those who actually performed the
experiments to ensure that
the MMBG paper was accurate.
As a
preliminary matter, I note that I have rejected any finding of fact
relating
to allegations concerning the Respondent's reporting of the
anesthesia used
in the Bengis experiments. The Respondent reported that only
sodium
pentobarbitone (also referred to as pentobarbital and nembutal) was
used as
an anaesthesia. Ex. 15. Dr. Bengis stated that he used both sodium
pentabarbital and xylazine, an analgesic. Agency Hearing Ex. C, p. 30; Tr.,
Jan 15, 1986, p. 12. The Respondent admitted that he had not remembered that
xylazine was also used until after the publication of his paper. Ex. 148, p.
3. Since the Agency did not dispute that xylazine is not an anaesthesia, but
was used only in conjunction with the anaesthesia, I have concluded that it
was not per se inaccurate, although incomplete, to report that only sodium
pentobarbitone was used.
I also make no finding that the
Respondent inaccurately reported that the
experimental procedures included
tests run by his laboratory for the
aldosterone levels in blood samples
obtained by Dr. Bengis. The key issue here
is whether the Respondent's
laboratory tested the blood samples for
aldosterone and renin levels, or
whether Dr. Cowley's laboratory tested the
samples for renin levels alone.
Although the evidence the Respondent presented
is not very persuasive, I
find that DHHS did not effectively rebut some of
that evidence. Thus, DHHS
did not meet its burden of proof to establish that
the Respondent's
laboratory did not test the blood samples. Specifically, DHHS
presented no
evidence from direct sources regarding the handling and testing
of blood
samples. While I find the testimony of DHHS's witnesses to be highly
credible, it is limited by the extent to which they had actual knowledge of
the occurrences and could remember the events in question. Although Drs.
Bengis and Coleman, the DHHS witnesses, testified that the blood samples
were
given to Dr. Cowley's laboratory, the Agency presented no witnesses
from that
laboratory who could verify that the samples were in fact tested
there or
describe the procedures used to test the samples. [FN8]
The testimony of Drs. Bengis and Coleman regarding blood samples was in
direct conflict with the testimony of the Respondent, Dr. Connie McCaa, and
the Respondent's two laboratory technicians. The latter witnesses all
testified that Dr. Bengis brought samples to the Respondent's laboratory and
that these samples were tested by them. [FN9] Extensive testimony was
presented on the issue of the type of test tubes used by Dr. Bengis and the
labeling method used. [FN10]
I did not find the testimony of the
two laboratory technicians to be very
reliable. This finding is based
primarily on my assessment of the demeanor of
the witnesses and the extent
of their personal knowledge. Their testimony at
the hearing, if not
demonstrably inconsistent with prior statements to the
UMMC investigatory
panel, as the Agency argued, did not appear to be clear and
forthright. The
testimony was based on memories of relatively insignificant
and infrequent
interactions with Dr. Bengis during a period when the
technicians were not
employees of the Respondent and were present occasionally
on a voluntary
basis. See Tr. Apr. 2, 1986, pp. 51-52, 94-96. Especially in
light of the
Respondent's statement that his laboratory was responsible for
running
hundreds of aldosterone assays each week, their testimony contained a
degree
of detail regarding events of nine years ago which I did not find
credible.
Ex. 155, p. 3. Discussions between the Respondent and the laboratory
technicians may have enhanced their memories beyond the limits of their
personal knowledge. For example, Ms. Crawford appeared to have had little
actual interaction with Dr. Bengis and was involved in only some of the
alleged tests on his samples. Tr. Apr. 2, 1986, pp. 52 and 69-73. Thus, I
find
her testimony unreliable regarding the performance of the tests and the
total
numbers of samples tested. Ms. Gray, on the other hand, was apparently
involved in testing a larger number of samples which she believed to be from
the Bengis experiments. But her statements before the UMMC were rather vague
and indicated confusion regarding the dates testing occurred and the manner
in
which results were reported. Ex. 70b, pp. 210-213, 216-217. In light of
these
statements, I can not give much weight to her later assurance
concerning the
dates involved and the identification of a large number of
plasma samples with
the Bengis experiments.
The Respondent's
testimony, with respect to the receipt of blood samples and
subsequent
laboratory testing of the samples, was also of limited value. He
had little
direct involvement in most of the laboratory procedures and could
not speak
with any significant personal knowledge. These tests were actually
performed
by his technicians. Furthermore, the Respondent admitted that there
were
blood samples which could have been labeled in the same manner as the
Respondent asserted that Dr. Bengis' samples were labeled, from the
Respondent's experiments with captopril using sodium-depleted dogs. Ex. 148,
p. 4. I will discuss later the Respondent's testimony regarding the analysis
of the alleged laboratory test results, when relevant to issues related to
the
reporting of results. See Finding of Fact No. 11.
The
testimony of Dr. Connie McCaa was highly credible and supported the
Respondent's contention that rat blood samples from the Bengis experiments
were tested in the Respondent's laboratory. Dr. Connie McCaa's credibility
was
never challenged. She testified that the Respondent and his laboratory
technicians consulted with her regarding the problems involved in
aldosterone
assays on the small samples obtained from rat subjects in the
Bengis
experiments. While her testimony was credible, it was limited in
value because
she had little direct basis for her statement that the samples
involved were
from the Bengis experiments. Her understanding that she was
working on the
Bengis samples was apparently based on what others told her,
rather than any
direct knowledge of whose samples she worked with. Tr. Apr.
2, 1986, pp. 131-
32.
Although the evidence presented by the
Respondent was not very reliable,
cumulatively that evidence was sufficient
to place a burden upon DHHS to rebut
the Respondent's evidence with
affirmative evidence regarding the custody of
Dr. Bengis' blood samples. The
evidence presented by DHHS did not fully rebut
the Respondent's contention
that his laboratory tested blood samples from Dr.
Bengis' experiments for
aldosterone and renin. Although Dr. Bengis stated that
the blood samples
were tested in Dr. Cowley's laboratory, his memory of the
laboratory routing
of samples and of laboratory procedures was limited. Tr.
Jan. 15, 1986, p.
11; Ex. 160, p. 11; Tr. Apr. 3, 1986, p. 27-31. Furthermore,
Dr. Bengis was
uncertain what had been done with any blood left after testing
of the rat
blood samples had been performed, by his account, by Dr. Cowley's
laboratory. Ex. 160, p. 11; Tr. Jan. 15, 1986, p. 30. In sum, Dr. Bengis'
testimony did not directly address the routing of the blood
samples.
The other evidence presented by DHHS does not fully rebut the
testimony of
the laboratory technicians that they, in fact, tested a
substantial number of
samples, or the testimony of Dr. Connie McCaa that she
provided consultation
on the testing of Dr. Bengis' small rat blood samples.
DHHS presented no
affirmative evidence regarding the location of the blood
samples from Dr.
Bengis' experiments, the testing of those samples by
another laboratory, or
convincing evidence of other rat experiments at the
time which would have
accounted for the laboratory tests described by the
Respondent's witnesses.
DHHS presented no evidence that access to the blood
samples was restricted
after the samples had been refrigerated on the sixth
floor nor did DHHS
present a witness from Dr. Cowley's laboratory to testify
that the samples
were actually tested there and how they were disposed
of.
There is uncontested evidence that laboratory tests were performed
on rat
blood samples from experiments conducted by Ms. Henderson, another
graduate
student. But, while the Respondent initially stated that Ms.
Henderson's
samples from normal and sodium-depleted rats treated with
captopril might have
been confused, the Respondent ultimately did not rely
on this occurrence. In
any case, Ms. Henderson's principal experiments
involved a different protocol
and there is no evidence that her data from
hypophysectomized rats could have
produced results similar to those
reported. Similarly, while Dr. Bengis
testified that he was in the
Respondent's laboratory at other times during the
six years he worked at
UMMC, there is no evidence of other experiments
involving rat blood samples
which the Respondent's witnesses could have
confused in their memory with
the experiments during this time period.
The testimony of the
laboratory technicians and Dr. Connie McCaa was
sufficient to place a burden
on DHHS to further develop the chain of custody
of blood samples. In the
absence of any evidence that the samples were
actually received and tests
run by another laboratory, I must conclude that
the Respondent may have
obtained the samples, by some means, even if DHHS'
witnesses accurately
stated that the blood samples were given to another
laboratory. See also,
Finding of Fact No. 11.
a. The Respondent's paper has a notation in
the legend to the charts
illustrating the text which reads "N=12." The
finding of the NHLBI panel,
which was not contested by the Respondent, was
that this notation normally
indicates that each data point represents the
findings from 12 subjects. With
the number of data points presented, this
would require at least 360 test
subjects.
Dr. Bengis' thesis
clearly documents the number of rats actually used. The
generally high level
of detail in the thesis makes it a highly credible
source. Furthermore, the
thesis is a contemporaneous account of the Bengis
experiments, was not
written under any influence from these proceedings, and
was subject to
review by Dr. Bengis' advisory committee. According to the
thesis, a total
of 128 rats were used in the experiments. 42 rats were used
for the sodium
deficient group; the other 5 groups had, at most, 20 rats. Only
one
experimental group--the control group--had 12 rats. [FN11]
Each data
point represented an even smaller number of test subjects, since
the
collection of a blood sample was performed by decapitating the rat. Thus,
each test subject could only be used to produce one blood sample. For there
to
have been 12 test subjects for each data point, nearly all the subjects
in a
typical experimental group would have been used. If this had been true,
at
most one data point could have been obtained from each group, instead of
the
several which were reported.
The Respondent explained that the
apparent discrepancy arose out of a
communication problem with the publisher
because he believed the publisher was
requesting the number in each
experimental group, rather than the number for
each data point. Ex. 20, p.
4; Ex. 146, p. 4. This explanation does not change
the fact that the
notation inaccurately reported the number of test subjects,
as it would be
understood by an ordinary reader. Tr. Jan. 14, 1986, p. 62. It
merely
explains the process by which the inaccuracy occurred. Furthermore,
this
explanation does not address the fact that the number used by the
Respondent
did not correctly report either the number of subjects for each
data point
or the number of subjects in each group. The number was accurate
for the
total population of only one of the six groups. [FN12]
The use of N=12
was also part of a larger problem: the Respondent apparently
had no precise
knowledge of the size of the experimental groups or the fact
that the sizes
of the groups varied considerably. Neither the charts nor the
text of the
Respondent's article indicate that the data points represented
different
numbers of subjects. As discussed more fully below, I found highly
credible
the evidence presented by DHHS that the experimental procedures
involved the
sacrifice of half of each experimental group during the initial
control
phase. The inaccurate reporting of N=12 suggests that each data point
was
equally supported by data. Even if I were to accept that all of the data
points were supported by some data, not all of the data points could have
been
supported by results from the same number of subjects.
The
existence of this error is rendered even more egregious if I accept the
Respondent's testimony that he had with him what he believed were originals
or
copies of Dr. Bengis' raw data, either in the form of xerox sheets or a
bound
data book. This data would have clearly indicated the number of
subjects
involved.
The Respondent's explanation that the
inaccuracy was due to a communication
problem is not credible even when the
Respondent's rambling and imprecise
speaking style is considered. The
Respondent's explanation indicated generally
that the Respondent did not
have any precise personal knowledge of the actual
numbers of test subjects
used in the Bengis experiments. If he did, he could
have been able to give
more precise figures and would have responded to the
publisher in such a way
as to eliminate any misunderstanding by noting the
differing sizes of the
experimental groups. The Respondent's explanation also
indicated that the
Respondent either did not understand or simply disregarded
the importance of
precise reporting of experimental procedures.
b. The Respondent
reported that the test subjects had been kept for a two-
week control period
prior to captopril administration. Drs. Bengis and
Coleman, in their work
and in Dr. Bengis' testimony, reported that the test
subjects were kept for
only 3 or 4 days in a control period. Agency Hearing
Ex. C p. 38- 43; Tr.
Apr. 3, 1986, pp. 40-42; Tr. Jan. 15, 1986, pp. 209-10.
The Respondent
explained the discrepancy differently at the hearing and in
his post-hearing
brief. During the hearing, he stated that there was a two-
week "observation
period" during which the rats were monitored to determine if
they had
developed benign or malignant hypertension as a result of surgical or
dietary restrictions and to allow their blood pressures to stabilize. The
Respondent stated that this "observation period" occurred after implantation
of aortic catheters. Tr. Apr. 1, 1986, pp. 200-203. While the Respondent
explained that he considered this entire "observation period" to be a
control
period, he recognized that, according to their published works, Drs.
Bengis
and Coleman had only considered the last 3 or 4 days of the longer
"observation period" to be a source for control period data, since data was
only taken during that period, and the rats used as control samples in each
group were sacrificed. The Respondent typified the discrepancy as a mere
semantic misunderstanding as to whether the control period should be
considered only the period when control data was collected, or the entire
observation period prior to administration of captopril.
This
explanation, however, does not conform to the description of the
experiment
in the published work of Drs. Bengis and Coleman, and is explicitly
refuted
in Dr. Bengis' testimony. I find that the Respondent's explanation
does not
credibly rebut that evidence. Dr. Bengis' thesis is quite precise on
the
time periods involved at each stage of the experiment. In his thesis, Dr.
Bengis makes no reference to a two-week "observation period." Agency Hearing
Ex. C, pp. 30-44. In his testimony, Dr. Bengis explained why such a
prolonged
observation period would have been impractical: he stated that the
aortic
catheters placed in the rats would close up if left implanted for too
long a
time. This would result in unwanted rat deaths. Tr. Apr. 3, 1986, pp.
42-43.
See also, Tr. Jan. 15, 1986, p. 52.
In the Respondent's
post-hearing brief, he advanced another theory to
explain the reported
control period. The Respondent stated that the reported
control period
represented a two-week period in which parallel groups of test
subjects were
sacrificed at three or four weeks after surgery, to determine
whether the
aortic catheters could be inserted and captopril administration
started in
the primary group. Not only is this explanation inconsistent with
the
Respondent's own testimony that the reported control period was after the
aortic catheters had been inserted, it is also inconsistent with Dr. Bengis'
description in his thesis of the procedures used. The Bengis thesis reports
very specifically on the use of "parallel groups" in the experiment to
obtain
control phase data, and reports very specifically on the timing of
the samples
taken during the control phases. See, Ex. C, pp. 38-43. These
"parallel
groups" were not used in the sense that the Respondent suggests:
rather than
groups used for experimental purposes outside of the scope of
the reported
work of Drs. Bengis and Coleman, the evidence indicates that
the "parallel
groups" were used within the reported scope. These "groups"
provided control
phase data or ensured reserves to replace subjects which
died prior to the
expected time. Tr. Apr. 3, 1986, pp. 34-35; Tr. Jan. 16,
1986, pp. 55-58.
In the Respondent's comments on the NHLBI report, he
made a statement which
I consider to be a more probable explanation of the
inaccurate reporting. "Dr.
Bengis specifically stated that he used this two
week period of observation to
determine when to insert the aortic catheter
and begin treatment with
captopril. When I wrote the Methods and Results
sections of the manuscript in
London, I assumed the blood samples had been
collected during this two week
period of time." Ex. 148, p. 3 (emphasis
added). This explanation is
consistent with the Respondent's pattern,
discussed further below, of
reporting information based on his assumptions
or expectations rather than
empirical facts.
c. The Respondent
reported data on plasma renin activity (PRA) and plasma
aldosterone
concentration (PAC) in test subject rats during the period when
these rats
were being administered captopril. Drs. Bengis and Coleman, in
their
testimony during the hearings, clearly and emphatically stated that no
blood
samples were taken during this phase of the experiment. Tr. Jan. 15,
1986,
pp. 9-10, 17, 202, 209; Tr. Apr. 3, 1986, pp. 6-7, 11-12. This testimony
was
very credible and was never directly rebutted.
The Respondent's only
responses to this testimony were his general
allegations regarding parallel
studies and evidence that the protocol
initially developed by Drs. Coleman
and Bengis changed from the plans
originally made in early 1977. There is no
evidence that Dr. Bengis performed
any "parallel" rat studies similar to
those reported in his thesis, but
unreported there. Dr. Bengis specifically
denied that any parallel studies had
been performed. Tr. Apr. 3, 1986, pp.
34-36. The Respondent provided evidence
that the protocol planned by Drs.
Coleman and Bengis in early 1977 was
different in some key aspects from the
protocol which Drs. Bengis and Coleman
claimed to have actually followed.
See Respondent Hearing Exs. H and M. But
Dr. Coleman stated that they
abandoned this protocol before having advanced
beyond the preliminary stages
of the experiment, involving only normal and,
possibly, sodium deficient
rats. See Tr. Apr. 3, 1986, pp. 129-134. He stated
that the protocol had
been produced merely as a formality to maintain the
goodwill of the supplier
of the captopril, not as a record of the actual
events. Id., p. 134. He
further explained that the protocol had been modified
because the earlier
protocols had been overly complex, and because there were
doubts about the
technical feasibility of certain parts of the protocol, such
as the ability
to obtain accurate measurements in tests for renin when
captopril was also
present. Id., p. 153; Tr. Jan. 15, 1986, pp. 202-03.
Although the
Respondent claimed to have had data which he stated he believed
to be data
from blood samples taken during this phase of the experiment, he
testified
that he obtained his information about which data came from subjects
in
different experimental groups or at different phases in the experiment from
inferences drawn from test tube labels rather than any direct personal
knowledge of the experimental procedures. The Respondent's testimony was not
supported by any documentary evidence of the type of test tube labelling
used.
Because of the unreliability of the Respondent's source, I do not find
that
his testimony is sufficiently credible to establish that blood samples
were
taken during the period of captopril administration.
I also
find it relevant that the procedure was clearly set out in the
written works
of Drs. Bengis and Coleman, and did not include blood samples
taken during
the captopril stage. In each of six groups, rats were sacrificed
for blood
samples at only two points: before captopril administration, to
establish
control values, and after captopril administration, to establish
recovery
values.
The Respondent stated that he did not read Dr. Bengis' thesis
before writing
his paper. Ex. 146, p. 3. However, both the thesis and the
first article
written about the Bengis experiments were cited and
incorporated by reference
by the Respondent in his paper. Ex. 15, p. 70.
That contemporaneous act
indicated that he, at least, was willing to be held
responsible for full
knowledge of those works. Accepting that as an
admission, I conclude that if
the Respondent had substantively disagreed
with the procedures outlined in the
referenced works, or if additional
procedures had been performed to obtain
additional data, the Respondent
would logically have clearly stated his
disagreement and listed, as such,
any additional procedures in his paper.
The Respondent alleged,
generally, that he believed that the additional data
could have come from
"parallel studies" performed by Dr. Bengis at the same
time as the study
reported in Dr. Bengis' thesis. He offered no support for
this allegation,
which was specifically denied by Dr. Bengis. Dr. Bengis
stated that all of
his research with respect to related studies was described
in his thesis.
Tr. Apr. 3, 1986, pp. 34-35. Although Dr. Coleman referred to
the use of
parallel or companion groups of test subjects, he used the terms in
a
different sense. In the context of his testimony, it is clear that Dr.
Coleman was not referring to any separate groups of animals, merely the
control subjects in each experimental group. See Tr. Jan. 15, 1986, pp.
200-
202; Tr. Apr. 3, 1986, pp. 155-157. Dr. Coleman confirmed that no
additional
data had been generated by Dr. Bengis. Tr. Jan. 15, 1986, p. 212.
He explained
that one reason he and Bengis had decided not to test blood
samples from
subjects during the period of captopril administration was that
they believed
that captopril interfered with the accuracy of blood chemistry
tests. Tr. Jan.
15, 1986, pp. 202-203; Tr. Apr. 3, 1986, p. 153. He stated
that, instead, they
had only measured the characteristics of arterial blood
pressure and heart
rate. Id. With respect to "parallel groups," Dr. Coleman
stated that the only
such groups were those described in Dr. Bengis' thesis
to obtain blood samples
during the control period. These rats were treated
identically to the rats
given captopril, but were sacrificed for blood
samples prior to captopril
administration. Tr. Jan. 16, 1986, pp. 56-58; Tr.
Apr. 3, 1986, p. 145.
RESULTS REPORTED IN THE MMBG PAPER
10. THE SECTIONS WRITTEN BY THE RESPONDENT DID NOT ACCURATELY REPORT
THE
ACTUAL DATA RESULTING FROM THE BENGIS EXPERIMENTS.
11. THE
SECTIONS WRITTEN BY THE RESPONDENT REPORTED DATA FOR ANALYSES OF
BLOOD
SAMPLES TAKEN DURING THE PERIOD OF CAPTOPRIL ADMINISTRATION WHICH DID
NOT
RESULT FROM ACTUAL TESTS ON ANY SUCH BLOOD SAMPLES.
12. DHHS DID NOT
PROVE, BY EITHER A PREPONDERANCE OF THE EVIDENCE OR BY
CLEAR AND CONVINCING
EVIDENCE, THAT VALUES REPORTED BY THE RESPONDENT FOR
PLASMA ALDOSTERONE
CONCENTRATIONS DURING THE CONTROL PERIOD AND SUBSEQUENT TO
CAPTOPRIL
ADMINISTRATION COULD NOT HAVE RESULTED FROM THE BENGIS EXPERIMENTS.
Discussion of the Results Reported in the MMBG Paper
10. The
sections of this paper written by the Respondent report values for
data
which vary, in some instances quite significantly, from the values
reported
in the accounts of the Bengis experiments written by Drs. Bengis and
Coleman. With respect to this point, I consider only the data in this paper
for which values were also reported by Drs. Bengis and Coleman. After an
examination of the written descriptions of the experiments and the testimony
presented, I conclude that the Respondent inaccurately reported the results
of
the Bengis experiments.
I note specifically my disagreement
with the Respondent's assertion that the
results reported by the Respondent
were "almost identical" with the results
reported by Drs. Bengis and
Coleman. Respondent's Post-Hearing Brief, p. 76.
My review of the
comparative chart submitted by the Respondent indicates
variations, between
the results reported by the Respondent and the results
reported by Drs.
Bengis and Coleman, which appear too large to be attributable
only to
rounding errors or the differences in averaging techniques. The
Respondent
did not explain these discrepancies.
I do not reach any finding on the
actual source of the data in the
Respondent's paper. The Respondent claimed,
during the hearing, to have been
in possession of Dr. Bengis' original data
book and test results, but did not
explain why his reported data varied from
other reported data, particularly
the data reported by Dr. Bengis himself in
his thesis. Furthermore, the
Respondent's testimony was inconsistent with
his earlier statement to the
NHLBI panel in which he said that he did not
have Dr. Bengis' original, bound
data book, but only "xerox sheets of all
his data in renovascular hypertensive
rat models." Ex. 155, p. 4.
Nevertheless, I agree with the Respondent that it
is possible that he had at
least some original Bengis data. Certainly, the
Respondent had data provided
to him for publication in abstracts.
The evidence presented by DHHS,
regarding the custody of Dr. Bengis'
original data, is unclear. Dr. Bengis
was unsure concerning what he did with
his notebook containing data. He did
not believe that he left the notebook
with the Respondent, but was unsure if
he had destroyed it or left it with Dr.
Coleman. Tr. Apr. 3, 1986, pp.
30-31. A letter from Dr. Bengis to Dr. Coleman,
written shortly after Dr.
Bengis had returned to South Africa in 1978,
indicated both that Dr. Bengis'
original data had been left with Dr. Coleman
and that some of the data was
possibly "lost" before August 11, 1978. Ex. 13.
In addition, Dr. Connie
McCaa testified that she saw the Respondent with a
notebook he identified as
Dr. Bengis' notebook, while he prepared the MMBG
paper in the fall of 1978.
Tr. Apr. 2, 1986, pp. 133-134, 137-138. The lack of
clear evidence
concerning the notebook leads me to conclude that the
Respondent could have
had either Dr. Bengis' notebook, copies of the data, or
even a notebook of
some sort which he himself had prepared from another source
of the raw data.
Tr. Jan. 17, 1986, pp. 87-89.
Whether or not the Respondent had all of
Dr. Bengis' original data, the
Respondent provided no satisfactory
explanation for the discrepancies between
the Respondent's reports of the
data and other reports. I can only conclude
that, if the Respondent had
original data, the Respondent either did not have
a complete set of that
data and inaccurately reported complete results based
on such partial data,
or that the Respondent had complete results but reported
these inaccurately,
perhaps because he had insufficient knowledge of the
experimental protocol
or of Dr. Bengis' method of recording the results of his
experiments.
Whether or not the Respondent had Dr. Bengis' data, my conclusion
remains
that the Respondent inaccurately reported the data.
In particular, my
conclusion is based on the values reported by the
Respondent for: a) the
arterial pressure of normal rats and salt-deficient
rats during the control
period; b) the PRA for normal rats after captopril
administration c) the
arterial pressure for salt-deficient rats after
captopril administration; d)
the arterial pressure for two-kidney Goldblatt
benign rats after captopril
administration; e) the arterial pressure for two-
kidney Goldblatt malignant
rats during the control period; and f) the PRA for
one- kidney Goldblatt
malignant rats sacrificed during the control period.
These values differed
widely, on their face, from other reports of data
resulting from the Bengis
experiments. Differences in other values, while
present, did not vary enough
for me to conclude that difference was
significant without more evidence to
persuade me of the degree of precision
customary in scientific
reporting.
The Respondent suggested that results may differ because of
a different
selection of the test subjects reported in the results. He
stated that, in
reporting results, a scientist may omit test subjects with
seemingly
inconsistent or incorrect results. While this may be proper, the
values
reported by the Respondent are sufficiently out of the range of those
reported
by Drs. Bengis and Coleman that I find this after-the-fact
explanation to be
lacking in credibility. [FN13] Furthermore, Drs. Bengis
and Coleman testified
that they reported all the results obtained, and did
not eliminate any test
subjects. The Respondent has not alleged that he
eliminated any test subjects
from his report. Particularly when no
indication was made that a smaller
sample size was used (the notation N=12
would, as discussed earlier, even
suggest a larger sample size), and when
the published findings of Drs. Bengis
and Coleman were apparently known to
the Respondent, I find no evidence to
support the Respondent's
suggestion.
Similarly, there is no evidence to support an explanation
based upon any
type of mere mathematical error. The errors do not appear to
follow any
pattern or to be predictable in any manner.
11. The
Respondent reported data for analyses of blood samples taken during
the
period of captopril administration. None of the other reports of the
Bengis
experiment mention that blood samples were taken during that period. As
I
discussed earlier, in Finding of Fact No. 9(c) concerning the Respondent's
general inaccuracy in reporting that blood samples were taken during that
period, I find these other reports, and the testimony of Drs. Bengis and
Coleman, to be persuasive.
I, therefore, conclude that for this
data the Respondent invented, or
fabricated, values based upon his extensive
knowledge and understanding of the
field. The record shows that the
Respondent has considerable expertise in his
field. He has sufficient
knowledge to examine partial data, or data for only
certain biological
characteristics, such as arterial pressure and blood
pressure, and determine
what the data might reasonably be for other values not
actually measured.
His extensive experience in experiments with dogs given
captopril and
similar substances would have shown him the probable parameters
of results
for rats under the protocol of the Bengis experiments. See, e.g.,
Ex. 11;
Respondent Hearing Ex. C.
I discussed above the possibility that the
reported results were the result
of some confusion of Dr. Bengis' samples
with other samples being processed by
the Respondent's laboratory. The
Respondent did not rely on any such confusion
as a defense, and presented no
concrete evidence about particular samples that
could have been confused
with Bengis' samples and produced the reported
results.
This
finding includes findings with respect to aldosterone data during the
period
of captopril administration. I address below, at point 12, issues
raised by
the Respondent with respect to the allegations regarding other
alleged
fabrication of aldosterone data.
12. The Respondent reported values
for plasma aldosterone concentration
(PAC) tests for blood samples taken
from test subject rats. The Respondent
claimed that these values were
obtained in his laboratory. Drs. Coleman and
Bengis, in their written
accounts of the experiments and in their testimony,
stated that the
Respondent could not have obtained those values because: a)
blood samples
from these experiments were not given to the Respondent's
laboratory for
analysis; and b) the procedures for their experiment did not
include testing
for PAC.
I have already discussed the evidence with respect to the
procedures in the
stages of the experiment involving testing of blood
samples. I found that the
evidence was not sufficient to find that no blood
samples were tested in the
Respondent's laboratory.
In that
earlier discussion, I did not detail additional points made by the
Respondent with respect to aldosterone tests, because I did not find them
credible. I did not find credible the Respondent's suggestion that he could
not have fabricated aldosterone data because he did not know what
aldosterone
data would be in rats. The Respondent stated that he found the
aldosterone
data to be surprising and that he visited Jurg Muller's
laboratory in Zurich
in order to ask if the data was consistent with other
rat aldosterone data.
The assertion that he was ignorant of the expected
range of values is
inconsistent with the publication, in 1974, of a study in
which the Respondent
measured aldosterone levels in anaesthetized and
conscious rats. Respondent
Hearing Ex. C. Furthermore, the Respondent's
laboratory was testing
aldosterone levels in hypophysectomised as well as
for normal and sodium
depleted rats for the work of Bennye Henderson. The
values obtained from that
work may have resembled the values which could
have been obtained from rats in
the Bengis experiment. See Tr. Jan. 17,
1986, pp. 71-72. Considering the
Respondent's experience, I must conclude
that the Respondent would have known
the range of aldosterone values
normally found in a conscious rat. In
addition, I did not find credible the
testimony of the Respondent's witnesses
regarding the performance of PAC
tests.
As I stated earlier, however, in the absence of any evidence
establishing
that the samples were elsewhere, I find credible the testimony
of the
Respondent's witnesses regarding the presence of some of Dr. Bengis'
samples
in the Respondent's laboratory, even if I did not find their
testimony to be
credible regarding the number or experimental source of the
samples. The
testimony of the laboratory technicians and of Dr. Connie McCaa
that samples
were tested in the laboratory, and that Dr. Connie McCaa was
consulted on the
practical problems of performing aldosterone assays on the
small samples
resulting from the Bengis experiments, placed at least some
burden on DHHS to
produce rebuttal evidence. I find that DHHS did not
adequately rebut this
testimony to establish that the Respondent's
laboratory did not test blood
samples from the Bengis experiments.
Because I find that DHHS did not establish that the Respondent's laboratory
did not obtain blood samples from the Bengis experiments, I must conclude
that
DHHS did not prove that the Respondent's PAC data could not have been
based on
actual tests on blood samples from the Bengis
experiments.
Although Drs. Bengis and Coleman, who were in control of
other phases of the
experiment, did not include PAC analysis in the protocol
they ultimately
followed, I find that these tests could have been performed
by the Respondent.
The Respondent clearly had an interest in PAC analysis
and would logically
have performed such analysis if his laboratory tested
the blood samples.
Furthermore, the Respondent produced evidence that Drs.
Bengis and Coleman had
planned to have PAC analysis included in the study in
early 1977. Respondent
Hearing Exs. H and M. Although Drs. Bengis and
Coleman asserted that this
protocol had been changed, there was no
documentation of the timing of this
change which would rule out the
possibility that some, or even all, of the
samples were still subjected to
PAC analysis. The Respondent presented
evidence that sufficient blood was in
each sample to perform both the PRA test
required by Drs. Bengis and Coleman
and the aldosterone assays. If the
Respondent actually had blood samples and
was not aware, as he stated, that
the procedures required by Drs. Bengis and
Coleman no longer involved
adosterone assays, I agree that the Respondent
could have obtained some
aldosterone data. [FN14]
Although I find
that DHHS did not establish that the Respondent's laboratory
did not obtain
blood samples in some manner, I note that this finding is based
not on any
credible substantive presentation by the Respondent concerning the
manner in
which his laboratory obtained blood samples from the Bengis
experiments. My
conclusion is based on my findings that DHHS did not meet its
burden of
proof on this one point. I found the DHHS witnesses to credibly
describe all
phases of the experiments in which they directly participated.
These
witnesses could not, however, testify with such personal knowledge
regarding
the actual laboratory testing. In the absence of any such testimony,
I will
accept the generally less credible testimony of the Respondent's
witnesses
that some samples from the Bengis experiments were tested in the
Respondent's laboratory.
THE RESPONDENT'S INTENT TO DECEIVE
13. THE RESPONDENT KNEW OR SHOULD HAVE KNOWN THAT HE DID NOT HAVE
SUFFICIENT
INFORMATION CONCERNING THE EXPERIMENTAL PROCEDURES OR RESULTS TO
ACCURATELY
REPORT THE METHODS OR RESULTS WHICH ARE SET FORTH IN THIS
PAPER.
14. THE INACCURATE PROCEDURAL DESCRIPTIONS, THE INACCURATE
DATA, AND THE
DATA UNSUPPORTED BY EXPERIMENTATION WERE REPORTED BY THE
RESPONDENT BECAUSE OF
A CARELESS DISREGARD FOR ACCURACY IN HIS PUBLISHED
REPORTS. WITH THE EXCEPTION
OF THE REPORTING OF THE LENGTH OF THE CONTROL
PERIOD AND NUMBER OF SUBJECTS
FOR EACH DATA POINT, THE INACCURATE REPORTING
WAS ALSO BECAUSE OF THE
RESPONDENT'S INTENT TO MISLEAD THE READER INTO A
BELIEF THAT THE RESPONDENT
ACTUALLY PERFORMED OR OVERSAW THE REPORTED
EXPERIMENTS AND OBTAINED THE
REPORTED RESULTS. THESE INACCURACIES WERE NOT
MERE OVERSTATEMENTS OR
ACCIDENTALLY IMPRECISE RENDERINGS OF THE UNDERLYING
EXPERIMENT.
Discussion of the Respondent's intent to deceive
13. Viewing the record as a whole, it is clear that the Respondent was
not
involved in the actual performance of the Bengis experiments and had
little
actual knowledge of the procedures or results. The Respondent was not
involved
in any official capacity in the Bengis experiments and was not on
Dr. Bengis'
advisory committee. Although he testified that he was often in
Dr. Bengis'
laboratory and knew exactly what was being done, his testimony
is inconsistent
with the Dr. Bengis' statement that contact between the two
was informal and
less substantive. Tr. Jan. 15, 1986, p. 15; Ex. 35, p. 3.
The Respondent's
testimony is also inconsistent with the fact that he
reported procedures
inaccurately and was unable to testify concerning
precise details of the
experiments.
In his description of how he
interpreted any results which he may have had,
the Respondent admitted that
he had no actual knowledge of what the data, that
he allegedly had,
represented. He stated that he knew the general procedures
of the
experiment, and that he had data marked in a coded format which he
interpreted to be consistent with his understanding of the procedures. It is
apparent that such a practice lends itself to error and is inherently
unreliable. The Respondent made no attempt to verify the accuracy of his
assumptions regarding the data with either Drs. Bengis and Coleman, who
actually ran the experiments.
I find that the Respondent was
unable to show that he had, or even that he
could have reasonably believed
that he had, sufficient knowledge of the
experiments to accurately report
the methods or results. The inaccurate
reporting of methods and results, in
itself, is prima facie evidence of his
lack of knowledge.
Dr.
Bengis' testimony that contacts with the Respondent were not unusually
frequent, and were more conversational than substantive, supports a finding
that the Respondent did not have sufficient knowledge of the experiments.
Tr.
Jan. 15, 1986, p. 15. Furthermore, the Respondent's description of his
haphazard method of interpreting data based on test tube labels instead of
participation in, or any direct involvement with, the experiments confirmed
that the inaccuracies were not accidental, but were the result of a willful
or
negligent failure to obtain minimally adequate information about, or
understanding of, the experimental activities before reporting the methods
and
results.
14. Considering my findings on the preceding issues,
I conclude that the
Respondent could only have presented this paper with an
intent to deceive the
reader. While certain of the inaccuracies may have
resulted merely from
negligent mistakes, others could only have been made
with the knowledge that
the reader would believe that the Respondent
actually had performed or
overseen the reported experiments and had obtained
certain results which were
not the same as the results actually obtained by
Drs. Bengis and Coleman. The
reader would either believe that the Respondent
had performed separate,
independent experiments, or was reporting
supplemental results.
In particular, I find that the inaccurate
representation of the number of
subjects or of the length of the control
period could have been made as a
result of lack of care and general
inattentiveness to detail, rather than any
specific intent to deceive. There
is no evidence to suggest that these
inaccuracies were, in themselves, of
major significance. If these inaccuracies
had occurred outside of a larger
context of inaccuracy, a controversy of this
magnitude might not have
emerged. And these inaccuracies were explained,
albeit not excused, by
factors which would suggest a careless disregard for
the importance of
accurate reporting of experimental data, but not necessarily
a specific
intent to deceive. The Respondent's description of where and how he
went
about writing the paper and other circumstances such as the lack of
communication between the various parties involved in the experiments, the
apparent changes in the planned protocol after early 1977, the general
atmosphere of mistrust which appears to have been present between the
Respondent and some other faculty members, the Respondent's admitted failure
to use the other accounts of the experiments, and the Respondent's admitted
reliance on less than fully knowledgeable sources (such as his laboratory
technician) for information about the experiments, suggest a careless
disregard for accuracy in reporting scientific results. These factors
persuade
me that the Respondent may not have had a specific intent to
deceive in
presenting these inaccuracies.
With respect to the
other inaccurate elements of the paper, and the general
context in which
these inaccuracies were present, I find that the evidence
indicates both a
lack of due care and a specific intent to deceive. These
other inaccurate
elements, such as the reporting of data not obtained in the
actual
experiments, could only have been made with knowledge that the data
were
inaccurate and would deceive the reader into believing that the
Respondent
had been actually involved in the performance of the reported
experiments
and obtained the reported results. Since I have found that the
Respondent
knew or should have known that he had insufficient knowledge of the
experimental procedures and results to support the reported data, I must
conclude that he knew or should have known that he was creating a false
impression.
My finding that Respondent's actions were intentional
is underscored by his
failure to provide Dr. Bengis with the paper for
review prior to its
publication as might ordinarily occur. (This is in
contrast to Dr. Coleman's
handling of the other article written and
published after Dr. Bengis' return
to South Africa, which was provided to
Dr. Bengis for comment prior to
publication. Tr. Apr. 3, 1986, p. 48.)
Indeed, Dr. Bengis indicated that he
had no knowledge of the paper's
existence until the allegations concerning
falsified and fabricated data
surfaced in 1982. Tr. Jan. 15, 1986, p. 9.
As stated by Dr. Coleman,
who had supervised the Bengis experiments, when he
read this paper, he
understood it to be asserting that new experiments had
been performed in
addition to the Bengis experiments. The presentation of
different methods
and results would naturally lead a reader to such an
assumption. Tr. Jan.
15, 1986, pp. 198-199. The reported methods and results
are sufficiently
unrelated to the experiments reported by Drs. Bengis and
Coleman that I find
that the Respondent must have intended to produce such a
response in a
reader.
C. FACTS RELATED TO THE PAPER "THE EFFECTS OF ANGIOTENSIN I--CONVERTING
ENZYME
INHIBITORS ON ARTERIAL BLOOD PRESSURE AND URINARY SODIUM
EXCRETION"
AUTHORSHIP AND GENERAL SCOPE
15. IN FEBRUARY 1978, THE RESPONDENT SUBMITTED FOR PUBLICATION IN
CIRCULATION RESEARCH A PAPER ENTITLED "THE EFFECTS OF ANGIOTENSIN
I--
CONVERTING ENZYME INHIBITORS ON ARTERIAL BLOOD PRESSURE AND URINARY
SODIUM
EXCRETION" (THE MHM PAPER). THE MHM PAPER WAS SUBSEQUENTLY PUBLISHED
IN VOLUME
43, NO. 1, SUPP. 1 OF CIRCULATION RESEARCH. THE LISTED AUTHORS
WERE R.E. MCCAA
(THE RESPONDENT), J.E. HALL, AND C.S. MCCAA. EX.
11.
16. THE RESPONDENT WROTE THE ENTIRE MHM PAPER. MOST OF THE RESULTS
REPORTED
IN THE PAPER WERE BASED UPON EXPERIMENTATION PERFORMED BY THE
RESPONDENT AND
NOT A SUBJECT OF THIS INQUIRY. AT ISSUE HERE ARE THE RESULTS
REPORTED WITH
RESPECT TO TESTS OF THE EFFECTIVE RENAL PLASMA FLOW (ERPF) AND
GLOMERULAR
FILTRATION RATE (GFR) IN SODIUM DEFICIENT DOGS BEFORE AND DURING
THE
ADMINISTRATION OF CAPTOPRIL. THESE RESULTS WERE BASED UPON EXPERIMENTS
PERFORMED BY DR. JOHN E. HALL. THE RESPONDENT PERFORMED NO INDEPENDENT TESTS
TO OBTAIN THE REPORTED RESULTS IN THIS SECTION.
17. THE DISPUTED
SECTION OF THE MHM PAPER PURPORTS TO REPORT THE RESULTS OF
TESTS FOR
ARTERIAL PRESSURE, ERPF, AND GFR ON FOUR CONSCIOUS SODIUM DEFICIENT
DOGS.
THESE RESULTS WERE PRESENTED IN THE GRAPHS LABELED FIGURE 4, AND WERE
SUMMARIZED IN THE ACCOMPANYING TEXT.
Discussion of Authorship and Scope
The facts related to authorship of this paper (the MHM Paper) were not
contested by the parties. The paper reported on several related experiments
performed on conscious sodium deficient dogs. The part of the paper at issue
purported to present the results of an experimental procedure performed by
Dr.
Hall during September and October of 1977. Below I discuss the contested
findings of fact, numbered 18-25.
BASIS FOR THE DISPUTED PART OF THE MHM PAPER
18. DR. HALL PERFORMED TESTS ON ONLY TWO SODIUM DEFICIENT
DOGS.
19. THE RESPONDENT INACCURATELY REPORTED THAT THE DATA RESULTED
FROM TESTS
ON FOUR DOGS.
Discussion of the Basis for the Disputed Part of the MHM Paper
18. I found persuasive the testimony of Dr. Hall with respect to the
number
of dogs tested to obtain data. Dr. Hall conducted the tests for ERPF
and GFR
supposedly reported in the disputed part of the paper. He drew blood
from the
subjects, tested the blood, and reported, in some form, the data to
the
Respondent. He had personal knowledge of the methodology used and the
results
obtained. Furthermore, he appeared to be a meticulous and careful
person,
whose general truthfulness was never strongly questioned.
The Respondent did not directly contradict Dr. Hall's testimony with respect
to the number of test subjects or the results obtained. The Respondent's
testimony that four dogs were available to be tested did not contradict Dr.
Hall's statement that only two dogs were actually tested. Although the
Respondent directly contradicted Dr. Hall's specific testimony that two of
the
dogs were not in a satisfactory condition to provide blood samples for
analysis due to a non-functioning catheter in one dog and an infected leg on
the other dog, I find Dr. Hall's testimony to be more credible. Dr. Hall had
greater personal knowledge of the circumstances of his experiments. Dr.
Hall's
explanation was simple and direct, and indicated a degree of concern
regarding
the scientific validity of the results. The Respondent could not
show that he
had actually examined the animals on any of the days in
question, and
responded with only general statements about the Respondent's
usual practices
in handling animal subjects. Even if I accepted the
Respondent's testimony as
generally true, that would not fully rebut Dr.
Hall's testimony as to the four
specific test subjects during the particular
time period of this experiment.
19. In light of the findings above, it
is clear that the Respondent
inaccurately reported the number of test
subjects in his paper.
REPORTED RESULTS AND STATISTICAL ANALYSIS IN THE MHM PAPER
20. THE RESPONDENT INACCURATELY REPORTED THE PROCEDURES FOLLOWED AND
VALUES
OF THE DATA OBTAINED. THE RESPONDENT REPRESENTED AS THE AVERAGE OF
VALUES
OBTAINED ON THE SAME DAY FROM DIFFERENT SUBJECTS A VALUE WHICH
ACTUALLY WAS
THE AVERAGE OF VALUES OBTAINED ON DIFFERENT DAYS.
21.
DR. HALL HAD INFORMED THE RESPONDENT ABOUT THE INACCURATE REPORTING OF
THE
NUMBER OF TEST SUBJECTS IN THE MHM PAPER PRIOR TO THE PUBLICATION OF THE
MHM
PAPER AND THE RESPONDENT HAD AMPLE TIME TO SUBMIT CORRECTIONS, EVEN IF
THERE
WAS A MISUNDERSTANDING OR MISTAKEN COMMUNICATION AT THE TIME THE
RESPONDENT
WROTE THE PAPER.
22. THE RESPONDENT INACCURATELY REPORTED THE
STATISTICAL SIGNIFICANCE OF THE
DATA PRESENTED. THE STATISTICAL ANALYSIS WAS
NOT CONSISTENT WITH THE NUMBER OF
TEST SUBJECTS AND MAGNIFIED THE
SIGNIFICANCE OF THE DATA.
Discussion of the Reported Results and Statistical Analysis in the MHM Paper
20. During the testimony of Dr. Hall, the Agency introduced into the
record
a document which Dr. Hall testified was a table of the raw data from
his tests
on the Respondent's dogs. Agency Ex. N. This data table was titled
"Compiled
Data for Na-Depleted Dogs: Effects of 14,225 on Renal Function and
Fluid
Volumes." Dr. Hall testified that he had given the original copy of
this
document to the Respondent prior to the time that the Respondent wrote
the MHM
paper.
The Respondent specifically denied that he had ever
received this data table
and testified that Dr. Hall had given to him a
graph, from which the graphs in
Figure 4 were derived. The Respondent stated
that the graph he received was
merely a roughly drawn version of the
published graph, with "some numbers
written in." Tr. Apr. 2, 1986, pp.
68-71.
In general, I find Dr. Hall's account of the form in which he
gave the data
to the Respondent to be convincing. Dr. Hall appeared to be a
meticulous and
well organized person who would have retained copies of any
presentation of
the data which he had prepared. The Respondent offered no
clear reason why Dr.
Hall would have been motivated to lie about the
occurrence, except for
generalized assertions of personal animosity between
the two and the fact that
Dr. Hall was a member of the UMMC faculty. These
were not sufficient to
disturb my sense of Dr. Hall's reliability.
Furthermore, Dr. Hall's unrebutted testimony was that it would be impossible
for the Respondent to calculate the statistical significance presented in
the
MHM paper based solely upon a graph. Tr. Jan. 16, 1986, p. 143. In
response,
the Respondent claimed that Dr. Hall had provided the statistical
analysis.
Tr. Apr. 2, 1986, pp. 72-74. I find Dr. Hall's testimony to be
convincing in
establishing that he had not prepared the statistical
analysis. Dr. Hall
stated that the presentation of statistical significance
in the MHM article
was inconsistent with his customary manner of reporting
statistical
significance. Tr. Apr. 3, 1986, pp. 104-106, 108. I, therefore,
concluded that
the Respondent had indeed prepared the statistical analysis
of the
significance of the data and must have had a full data sheet at that
time.
From the compiled data sheet, it is clear that the testing of
animals did
not occur as indicated in the graphs at Figure 4 in the MHM
paper. The data
sheet indicates that values for GFR were obtained on the
first day of
administration with captopril in both animals, on the fourth
day in one animal
and on the seventh day in a second animal. The graph in
Figure 4 indicates
values for the first day and the seventh day only. The
value for the seventh
day appears to be an average of the values in the data
sheet for the different
animals tested on the third and seventh days.
Similarly, the data sheet
indicates that ERPF was tested on the second day
of Captopril administration
in one animal, the third day in another animal
and the eighth day in both
animals. The graph in Figure 4 indicates values
for the second and eighth days
only. The value for the second day appears to
be an average of the values of
the data sheet for the different subjects
tested on the second and third days.
Compare Ex. 11 with Agency Hearing Ex.
N; See Tr. Jan. 16, 1986, pp. 90-95.
21. The Respondent agreed that he
had been informed of this error in the MHM
paper prior to publication. Tr.
Apr. 2, 1986, pp. 83-84. Although the
Respondent testified that he was only
informed of the error in the MHM paper
subsequent to the time that the
Respondent had sent in the "galley proofs" of
the paper, his testimony
indicated that there would have been ample time to
withdraw the paper, or,
at least to withdraw the disputed part of the article.
Tr. Apr. 2, 1986, p.
90.
Even if there had been some misunderstanding or mistaken
communication at
the time the Respondent received the data from Dr. Hall and
wrote the paper,
the evidence is clear that he was aware of the error prior
to publication.
22. The Agency presented extensive testimony and
analysis of the raw data
used by the Respondent, which indicated that the
statistical analysis of the
significance of the values reported in the MHM
paper was not accurate. See Ex.
60, Tab 1, pp. 19-20 and App. II; Tr. Jan.
16, 1986, pp. 85-90, 179-182.
Testimony by the Deputy Director of the NIH
Division of Extramural Affairs,
who was an active researcher for 12 years in
a field related to hypertension,
indicated that the significance of the
values obtained by Dr. Hall was fairly
low due to the small number of test
subjects. Tr. Jan. 16, 1986, p. 89. This
same testimony indicated that the
statistical significance reported by the
Respondent was very high. See also,
Tr. Jan. 16, 1986, p. 180.
The Respondent submitted some worksheets
which he stated were done by a
mathematician and which he alleged supported
his claims that the significance
of the data was high. No evidence was
introduced regarding the qualifications
of that mathematician or the method
he had used. The worksheets themselves do
not indicate the method used and
it is not even clear to what figure the
worksheets refer. I find the
evidence presented by the Agency to be more
persuasive, in that it was
supported by the testimony of witnesses with known
experience and other
qualifications. The Agency witnesses were not challenged
on this point by
the Respondent on cross examination. And the evidence these
witnesses
presented included discussion of the methodolgy and reasoning which
supported their conclusions.
As discussed earlier, in connection
with discussion of the form in which the
Respondent was given data by Dr.
Hall, I reject the Respondent's contention
that he did not prepare the
statistical analysis himself. Also, since I find
that the Respondent was
given a data sheet which accurately reported the
number of test subjects. I
cannot find that the mistakes on the statistical
analysis were connected to
some unintentional mistake in the number of test
subjects.
THE RESPONDENT'S INTENT TO MISLEAD
23. PRIOR TO PUBLICATION OF THE MHM PAPER, THE RESPONDENT DID NOT
ATTEMPT TO
CORRECT THE INACCURATE NUMBER OF TEST SUBJECTS, THE INACCURATE
VALUES, OR THE
INACCURATE STATISTICAL ANALYSIS OF THE SIGNIFICANCE OF THE
DATA.
24. THE RESPONDENT KNEW OR SHOULD HAVE KNOWN THAT THE NUMBER OF
SUBJECTS,
CERTAIN REPORTED VALUES, AND THE STATISTICAL ANALYSIS WERE NOT
ACCURATELY
REPORTED IN THE MHM PAPER.
25. THE RESPONDENT CAUSED
THE PUBLICATION OF THE INACCURATE DATA AND
ANALYSIS WITH AN INTENT TO
MISLEAD READERS INTO THE BELIEF THAT THE REPORTED
RESULTS AND SIGNIFICANCE
HAD ACTUALLY BEEN OBTAINED THROUGH EXPERIMENTATION.
THE INACCURATE REPORTING
WAS INTENTIONAL AND NOT MERELY DUE TO ACCIDENTALLY
IMPRECISE STATEMENTS AND
GRAPHS OR TO ANY REASONABLE MISTAKEN COMMUNICATION.
Discussion of the Respondent's Intent to Mislead
23. The Respondent admitted that he did not attempt to change or
withdraw
the paper prior to publication, even after he was informed of the
error with
regard to the number of dogs tested. Tr. Apr. 2, 1986, pp. 92-94.
The
Respondent may have discarded the data sheet after preparing the graphs
in
figure 4, and thus may not have been able to correct any mistakes at the
"galley proof" stage. But this problem could have been corrected by asking
Dr.
Hall for another copy of the data sheet. Although I recognize that
personal
animosities may have existed between the Respondent and Dr. Hall,
this cannot
excuse Respondent's failure to further communicate with Dr. Hall
or to assure
that the results were reported accurately. See Tr. Apr. 2,
1986, pp. 93-94.
24. As I discuss above, I find that the Respondent
received accurate data
concerning the number of test subjects and the values
obtained by Dr. Hall.
When preparing the graphs and statistical analysis, he
must have been aware
that he was not accurately presenting the
experimentation performed. Whether
or not Dr. Hall specifically informed the
Respondent about all of the errors
does not matter since the errors are
clear when comparing the MHM paper with
the compiled data sheet.
25. Considering my findings on the preceding issues, I conclude that the
Respondent could only have written, submitted, and failed to withdraw or
correct the MHM paper because of an intent to mislead readers into the
belief
that the reported results were accurate and more significant than
they
actually were. I have found that the Respondent must have known of the
correct
data, and I can find no reasonable explanation why he inaccurately
reported
the data except for the intent to make the data appear more
significant.
Furthermore, I do not believe that the inaccurate
reporting was due merely
to sloppiness or oversight. I agree with the
testimony of Dr. Roscoe that
certain aspects of the inaccurate reporting, in
particular the pooling of data
values obtained on different days, require a
"conscious effort" and reflect
intentional misrepresentation of the actual
data. Tr. Jan. 16, 1986, p. 94-97.
I did not find convincing the
Respondent's assertion that he did not intend
to publish inaccurate data,
which he supported with his testimony that he had
told Dr. Hall to make any
necessary changes in the paper. Tr. Jan. 17, 1986,
pp. 131-137. The
Respondent could not reasonably have intended for Dr. Hall to
make changes
in a paper on which the Respondent was first author, on which the
Respondent
had been the only person to contact the journal, and which the
Respondent
had seen through the galley proof stage. The Respondent's testimony
that he
had certain problems in his personal relationship with Dr. Hall, is
further
evidence that the Respondent did not reasonably intend Dr. Hall to
make such
changes. See Tr. Jan 17, 1986, pp. 133-135.
D. FACTS RELATED TO SUBSEQUENT PUBLICATIONS
26. THE RESPONDENT CAUSED TO BE REPUBLISHED DATA FROM THE MHM PAPER,
WHICH I
HAVE CONCLUDED THAT HE REPORTED INACCURATELY, IN "ALDOSTERONE AND
FLUID
BALANCE," MCCAA, IN ANGIOTENSIN CONVERTING ENZYME INHIBITORS,
HOROVITZ, ED.
(1981); Ex. 147.
27. THE RESPONDENT CAUSED TO BE
REPUBLISHED THIS SAME INACCURATE DATA FROM
THE MHM PAPER, WITH THE ADDED
STATEMENT THAT WHEN CAPTOPRIL ADMINISTRATION WAS
DISCONTINUED, ARTERIAL
BLOOD PRESSURE, ERPF, AND GFR ALL RETURNED TO CONTROL
VALUES, IN TWO PAPERS:
"ROLE OF THE RENIN-ANGIOTENSIN SYSTEM IN HYPERTENSION,"
MCCAA, IN
PROCEEDINGS OF SYMPOSIUM ON ANGIOTENSIN CONVERTING ENZYME INHIBITORS
HELD IN
PUERTO RICO (1980); "REGULATION OF SODIUM EXCRETION, RENAL FUNCTION
AND
ARTERIAL PRESSURE," MCCAA, IN THE ROLE OF SALT IN CARDIOVASCULAR
HYPERTENSION, FREGLY AND KARE, EDS. (1982).
28. THE RESPONDENT
WROTE TWO MANUSCRIPTS WHICH CONTAINED THIS INACCURATE
DATA FROM THE MHM
PAPER, WITH ADDITIONAL DATA INDICATING THE EFFECT OF
ANGIOTENSIN II INFUSION
DURING CAPTOPRIL ADMINISTRATION. ONE OF THESE PAPERS,
"ROLE OF THE
RENIN-ANGIOTENSIN AND KALLIKREIN-KININ SYSTEMS IN THE CONTROL OF
FLUID AND
ELECTROLYTE METABOLISM, RENAL FUNCTION AND ARTERIAL BLOOD PRESSURE,"
OLSEN
AND MCCAA, WAS WITHDRAWN FROM THE PUBLISHER AFTER HAVING BEEN ORALLY
PRESENTED AT THE SYMPOSIUM ON MINERALOCORTOIDS IN ESSENTIAL AND SECONDARY
HYPERTENSION IN 1982. THE OTHER PAPER, "HORMONAL CONTROL OF SODIUM EXCRETION
AND ARTERIAL PRESSURE," MCCAA AND OLSEN, WAS WITHDRAWN FROM PUBLICATION
PRIOR
TO ANY PUBLIC PRESENTATION.
29. THE RESPONDENT DID NOT
INDEPENDENTLY PERFORM ANY EXPERIMENTATION WHICH
RESULTED IN THE RESULTS
REPORTED IN THESE PAPERS. THE RESPONDENT AGREED THAT
THE PAPERS REPORTED ON
THE SAME EXPERIMENTS PERFORMED BY DR. HALL.
30. DR. HALL DID NOT
PERFORM EXPERIMENTS WHICH RESULTED IN THE ADDITIONAL
DATA
REPORTED.
31. THE ADDITIONAL STATEMENTS AND DATA WERE INACCURATE
BECAUSE THEY
REPRESENT THAT THE RESPONDENT HAD ACTUALLY PERFORMED
EXPERIMENTAL RESEARCH AND
OBTAINED THE RESULTS SET FORTH. THE ADDITIONAL
STATEMENTS AND DATA DO NOT
INDICATE THAT THE RESULTS ARE HYPOTHETICAL,
PRELIMINARY, OR MERELY LOGICAL
DEDUCTIONS BASED ON CERTAIN KNOWN
VARIABLES.
32. THE RESPONDENT CAUSED THE REPUBLICATION OF THE
INACCURATE DATA FROM THE
MHM PAPER, AND SOUGHT FURTHER REPUBLICATION, WITH
THE INTENT TO MISLEAD
READERS INTO A BELIEF THAT HE HAD CONDUCTED RESEARCH
AND OBTAINED THE RESULTS
REPORTED. THE RESPONDENT'S ACTIONS WERE NOT DUE TO
ANY MISUNDERSTANDING WITH
RESPECT TO THE ACCURACY OF THE DATA OR A MERELY
NEGLIGENT OVERSIGHT.
Discussion of Facts Related to Subsequent Publication
26. The Respondent did not contest his authorship of the paper
"Aldosterone
and Fluid Balance," or the fact that the paper contains the
data from the MHM
paper which I have determined was inaccurately reported.
Finding of Fact Nos.
20-22. The Respondent asserted that the publication of
the paper without
corrections was inadvertent. He stated that he had
intended to change the
paper when it was returned for his review, but he
stated that the paper had
not been returned for his review prior to
publication. This testimony was
directly contradicted by the testimony of
the editor of the book where this
paper appeared, who indicated that Dr.
McCaa did have the opportunity to
review the manuscript of his paper. Tr.
Apr. 3, 1986, pp. 72-74.
I do not find credible the Respondent's
explanation that the republication
of the data was inadvertent. The
republication was subsequent to the original
inaccurate publication, which I
found earlier was intentional. Finding of Fact
Nos. 23-25. The Respondent
never adequately explained why the inaccurate data
was included in the
manuscript. The Respondent admitted that he was aware the
statement N=4 was
inaccurate at the time he submitted the manuscript, because
he had been told
by Dr. Hall in relation to the MHM paper, and yet that
statement was still
in the manuscript. In stating that he had intended to
change the N=4 in the
final version, the Respondent did not even address the
inaccurate reporting
of the values for the data. Even if the N=4 had been
changed, these would
still have been inaccurate.
27. The Respondent did not contest his
authorship of the papers, "Role of
the Renin-Angiotensin System in
Hypertension," Agency Hearing Ex. H, and
"Regulation of Sodium Excretion,
Renal Function and Arterial Pressure," Ex.
18, or the fact that these papers
contain data from the MHM paper which I have
already determined was
inaccurately reported. Finding of Fact Nos. 20-22.
With respect to the
statement that when captopril administration was
discontinued, arterial
blood pressure, ERPF, and GFR all returned to control
values, the Respondent
admitted that he had not actually measured these
values. Respondent's
Post-Hearing Brief, p. 53. He asserted, however, that the
statement did not
refer to precise values and was intended to be a statement
of biological,
scientific fact rather than a report of experimental results.
The
Respondent could not, however, identify any publication of this
"biological fact" prior to the time these manuscripts were submitted for
publication. Although the Respondent cited works from which he could have
derived a general knowledge of the biological nature and operations of
captopril, he has not cited any statement of this precise fact. The
Respondent
may have extrapolated or surmised from the cited materials the
probable
results when captopril was discontinued, with his understanding of
the subject
matter. But that does not provide an adequate basis for a
statement which
implies that an actual observation was made in the course of
an experiment.
Although the Respondent did not include actual data
points for this
statement, the context suggests that the statement
represents an actual
observation. As the Respondent admitted, the statement
was not, in fact,
supported by any actual observations, nor was it
referenced by any citation to
a work which would support it and demonstrate
a source with an experimental
foundation.
Moreover, the Respondent
did not initially characterize this statement as a
"biological fact."
The Respondent's earlier response when this statement was
questioned was
that it "was based on scientific observations and on
assumptions....
Subsequent experimental research studies have proved [this
statement]." Ex.
86, p. 8. The Respondent also had stated earlier that, "[w]e
have since
performed these studies, 'after the fact', and ERPF and GFR do
indeed return
to the levels that existed in the pretreatment period." Ex. 146,
p. 12. The
Respondent's "biological fact" response becomes, in context, a
further
effort to excuse the reporting of experimental results not supported
by
actual observations. The testimony of one of the Respondent's witnesses,
Dr.
Zola Horowitz, supports a conclusion that this statement would not be an
acceptable way to treat a "biological fact." [FN15] Tr. Apr. 3, 1986, pp.
62-
64 and 80-83.
28. The Respondent did not contest his authorship
of the papers, "Role of
the Renin-Angiotensin and Kallikrein-Kinin Systems
in the Control of Fluid and
Electrolyte Metabolism, Renal Function and
Arterial Blood Pressure," Ex. 1,
and "Hormonal Control of Sodium Excretion
and Arterial Blood Pressure," Ex. 2,
both of which had Dr. Olsen listed as a
co-author. The Respondent admitted
that he wrote the papers, and did not
contest that the papers contained data
from the MHM paper which I have
determined was inaccurately reported. Finding
of Fact Nos. 20, 21, and
22.
With respect to the additional data indicating the effect of
angiotensin II
infusion during captopril administration, the Respondent
never identified a
source for this data. The data was presented in detailed
numerical values,
along with values for standard deviations indicating the
degree of precision.
This level of detail suggests an experimental basis for
the data, but the
Respondent admitted that the data was "postulated" and not
the result of any
actual experimentation. Respondent's Post-Hearing Brief,
p. 55. Although
similar studies were done by Dr. Hall, the values reported
were not the same
as those reported by Dr. Hall, and the work of Dr. Hall
was not referenced by
any citation.
The Respondent argued that he
never intended these manuscripts to be
published in the form in which they
are in the record. Although there is some
question regarding this point, I
accept the Respondent's contention that these
manuscripts were rough drafts
not intended for publication. This contention,
however, does not explain why
the Respondent would put inaccurate date into a
rough draft.
29.
The Respondent admitted that he did not conduct any experiments
independently to supplement those conducted by Dr. Hall. While the
Respondent
referred to additional studies by Dr. Hall in his testimony, the
papers and
manuscripts do not cite to these additional studies and do not
contain values
identical to those found by Dr. Hall. See Ex. 257.
30. The additional studies performed by Dr. Hall did not produce any of the
data values reported by the Respondent in any of these papers. Although the
additional studies did include experimentation with infusions of both
captopril and angiotensin II, these experiments resulted in different values
than those reported by the Respondent in his manuscripts. Compare Ex. 11 and
Ex. 257. The additional Hall studies did not produce any results indicating
a
return to control values after captopril administration had been
discontinued.
31. As I stated earlier, the additional statements and
data appear to
reflect the results of experiments actually conducted. No
modifiers or
disclaimers indicate that the data is hypothetical or
preliminary. No
citations to other works indicate a valid experimental basis
for the data.
32. Considering my findings on the preceding issues, I
conclude that the
Respondent had an intent to mislead readers into a belief
that the reported
results had actually been obtained through
experimentation. Although the
Respondent stated that the republication of
the inaccuracies was inadvertent,
I find that the intent to mislead implicit
in the original publication was
continued in all subsequent publications for
the reasons described in Finding
of Fact Nos. 23, 24, and 25. Cf. Tr. Apr.
2, 1986, pp. 95-96. Republication
with knowledge of the inaccuracy can not
be excused under these circumstances.
E. FACTS RELATED TO THE RESPONDENT'S GRANT APPLICATIONS
33. THE RESPONDENT SUBMITTED GRANT APPLICATION NO. HL-09921-13 TO
NHLBI FOR
RENEWAL OF HIS GRANT "MECHANISMS RELATED TO CONGESTIVE HEART
FAILURE" IN
JANUARY 1977. Ex. 263.
34. THE RESPONDENT REPRESENTED,
ON PAGES 97 AND 105 OF THAT APPLICATION,
THAT HE WOULD BE USING GRANT FUNDS
TO PERFORM EXPERIMENTS ON ONE-KIDNEY AND
TWO- KIDNEY GOLDBLATT HYPERTENSIVE
SUBJECTS.
35. THE RESPONDENT SUBMITTED GRANT APPLICATION NO.
HL-09921-12S2 TO NHLBI
FOR RENEWAL OF HIS GRANT "MECHANISMS RELATED TO
CONGESTIVE HEART FAILURE" IN
APRIL 1978. Ex. 265.
36. THE
RESPONDENT REPRESENTED, ON PAGE 3 OF THAT APPLICATION, THAT HE WOULD
BE
USING GRANT FUNDS TO PERFORM EXPERIMENTS ON "FOUR MODELS OF EXPERIMENTAL
RENOVASCULAR HYPERTENSION IN RATS", INCLUDING "ONE-KIDNEY AND TWO-KIDNEY
BENIGN AND ONE AND TWO KDINEY [SIC] MALIGNANT RENOVASCULAR HYPERTENSIVE
RATS."
37. THE RESPONDENT SUBMITTED GRANT APPLICATION NO. HL-09921-14
TO NHLBI FOR
RENEWAL OF HIS GRANT "MECHANISMS RELATED TO CONGESTIVE HEART
FAILURE" IN 1979.
Ex. 266.
38. THE RESPONDENT REPORTED, ON PAGE 5
OF THAT APPLICATION, THAT THE MMBG
AND THE MHM PAPERS HAD BEEN PUBLISHED OR
WERE "IN PRESS" AND HAD RECEIVED
SUPPORT FROM HIS GRANT.
39. THE
RESPONDENT REPRESENTED, ON PAGES 10 AND 12 OF THAT APPLICATION, THAT
HE HAD
PERFORMED THE EXPERIMENTS THAT HE REPORTED IN THE MMBG AND THE MHM
PAPERS.
40. THE RESPONDENT SUBMITTED GRANT APPLICATION NO.
HL-09921-15 TO NHLBI FOR
RENEWAL OF HIS GRANT "MECHANISMS RELATED TO
CONGESTIVE HEART FAILURE" IN MARCH
1980. Ex. 267.
41. THE
RESPONDENT REPORTED, ON PAGE 5 OF THAT APPLICATION, THAT THE MMBG
PAPER HAD
BEEN PUBLISHED AND HAD RECEIVED SUPPORT FROM HIS GRANT.
42. THE
RESPONDENT REPRESENTED, ON PAGES 10 AND 11 OF THAT APPLICATION, THAT
HE HAD
PERFORMED THE EXPERIMENTS THAT HE REPORTED IN THE MHM PAPERS, AND
REPORTED
THE RESULTS WHICH I HAVE CONCLUDED THAT HE REPORTED INACCURATELY.
THIS
PRESENTATION INCLUDED A GRAPH WHICH INACCURATELY REPORTED THAT THE NUMBER
OF
SUBJECTS INVOLVED IN THE EXPERIMENT WAS FOUR.
43. THE RESPONDENT
SUBMITTED GRANT APPLICATION NO. HL-09921-17 TO NHLBI FOR
RENEWAL OF HIS
GRANT "MECHANISMS RELATED TO CONGESTIVE HEART FAILURE" IN
OCTOBER 1981. Ex.
227.
44. ON PAGES 81-90 OF THE GRANT APPLICATION, THE RESPONDENT
INCLUDED THE
DATA FROM THE MMBG PAPER WHICH I HAVE CONCLUDED THAT HE
REPORTED INACCURATELY.
45. ON PAGE 76 OF THE GRANT APPLICATION, IN THE
FORM OF A GRAPH, THE
RESPONDENT PRESENTED THE DATA FROM THE MHM PAPER WHICH
I HAVE CONCLUDED THAT
HE REPORTED INACCURATELY. ALTHOUGH THE REFERENCE TO
THE NUMBER OF SUBJECTS WAS
OMITTED, THE DATA POINTS WERE
INACCURATE.
46. ON PAGE 76 OF THE GRANT APPLICATION, THE RESPONDENT
PRESENTED THE
ADDITIONAL STATEMENT THAT VALUES FOR ARTERIAL BLOOD PRESSURE,
ERPF, AND GFR
RETURNED TO PRETREATMENT LEVELS AFTER CAPTOPRIL ADMINISTRATION
WAS
DISCONTINUED WITH THE SUBJECTS OF THE MHM PAPER, WHICH I HAVE CONCLUDED
WAS
UNSUPPORTED BY EXPERIMENTAL DATA.
Discussion of the Facts Related to the Respondent's Grant Applications
I will not separately discuss here the underlying determinations that
certain reported material was inaccurate, since I have fully discussed these
determinations above. The Respondent did not contest the following facts: he
submitted the grant applications; the grant applications represented that
the
Respondent would perform the experiments described in the MMBG and MHM
papers
with support from his grant, and the grant applications contained
material
which I have determined to be inaccurately reported. The remaining
elements of
the findings in this section are supported by the reasoning of
the preceding
findings, and I will not repeat that discussion.
VI. Conclusions of Law
A. ORDINARY STANDARDS OF CONDUCT FOR RESEARCH
1. THE RESPONDENT DID NOT ADHERE TO CERTAIN STANDARDS OF CONDUCT FOR
REPORTING SCIENTIFIC RESEARCH WHICH WOULD BE EXERCISED BY AN ORDINARY
RESEARCH
SCIENTIST. THESE STANDARDS ARE:
a. REPORTING
EXPERIMENTAL PROCEDURES ACCURATELY;
b. REPORTING
EXPERIMENTAL RESULTS ACCURATELY;
c. REPORTING EXPERIMENTAL
RESULTS ONLY WHEN THERE IS SUPPORTING EMPIRICAL
DATA, WHEN THOSE RESULTS ARE
NOT IDENTIFIED AS HYPOTHETICAL OR PRELIMINARY.
2. THE RESPONDENT
REPORTED DATA AND DESCRIPTIONS OF EXPERIMENTAL PROCEDURES
WITHOUT SUFFICIENT
KNOWLEDGE OF THE ACTUAL EXPERIMENTAL DATA AND PROCEDURES TO
ENSURE THAT
ARTICLES PREPARED FOR PUBLICATION MET THESE STANDARDS OF CONDUCT
FOR
REPORTING SCIENTIFIC RESEARCH WHICH WOULD BE EXERCISED BY AN ORDINARY
RESEARCH SCIENTIST.
3. THE RESPONDENT'S DEVIATIONS FROM THESE
STANDARDS OF CONDUCT SHOW EITHER A
DISREGARD FOR OR A LACK OF KNOWLEDGE OF
GENERALLY ACCEPTED RESEARCH PRACTICES.
Discussion of the Ordinary Standards of Conduct for Research
1. Neither party in this case has submitted any authoritative text or
regulatory guidance which clearly indicates the standards of conduct which
would be exercised by an ordinary research scientist. In measuring the
Respondent's acts against such standards, I have relied on the oral
testimony
of the scientists appearing as witnesses at the hearing in this
case. Although
I became aware that such standards are continually evolving
and that there may
be some areas of regional variation or personal
discretion, I was able to
discern a few fundamental principles which seemed
common to the scientific
community as a whole. The standards relevant here
relate primarily to the
accurate reporting of experimental procedures and
data, sufficient to allow
critical evaluation and replication of the
experiment.
I find that the Respondent violated these standards of
conduct in the
publications considered in the findings of fact, as described
in detail below.
a. I found that the Respondent inaccurately described
the procedures used in
the experiments underlying both the MMBG and MHM
papers. Finding of Fact Nos.
9 and 20. Although few explicit statements
regarding standards of conduct are
in the record, testimony from several
witnesses established that accurate
reporting of procedures used in an
experiment, so that readers can critically
evaluate the methodology and
results, is essential to scientific reporting.
See Tr. Jan. 14, 1986, pp.
174-176, 201; Tr. Jan 15, 1986, pp. 61-63, 66, 71,
76, 87; Tr. Jan. 16,
1986, p. 91. If the procedures are accurately reported,
readers can
independently assess the validity of the experimental method and
attempt to
repeat the experiment to test the accuracy of the results and the
analysis.
Accurate reporting also permits readers to evaluate the overall
significance
of the reported experiment. This testimony was not disputed by
the
Respondent.
For example, in the MHM paper, the Respondent reported
that values had been
obtained on the same day for certain procedures, when
actually those values
were the average obtained on different days. Ex. 11,
p. I-36. Dr. Roscoe, a
former research scientist and currently a Deputy
Director of a division of the
NHLBI, testified that this pooling of data was
an acceptable scientific
practice only "if you say the data were pooled."
Tr. Jan. 16, 1986, p. 91.
Similarly, the inaccurate reporting of the number
of subjects in the MMBG
paper would cause a reader to accord more
significance to the findings
reported than was warranted by the actual
experimentation.
b. I found that the Respondent reported inaccurate
data in both the MMBG
paper and the MHM paper. Finding of Fact Nos. 10, 11,
and 20. These papers
purport to report the actual results of experiments
performed. Dr. Dustan,
whose testimony was not rebutted, stated that
accuracy of facts and data is
essential to scientific reporting. Tr. Jan.
15, 1986, pp. 62-64. While her
distinguished career indicates that her
standards may be higher than the
ordinary scientist, there seems little room
for compromising this simple
principle. Other evidence in the record
supports the finding of a standard of
conduct requiring accurate reporting
of experimental data. When data is
published as an experimental result, the
ordinary scientist assumes that the
result was actually obtained by
experimentation. Id.; see, also, id., pp. 172-
173. In the past, when the
scientific community has discovered instances when
data unsupported by
experimentation was published as the results of an actual
experiment, public
censure has been invoked and the publications retracted.
Id., pp.
68-69.
c. I found that the Respondent reported data which he claimed
were intended
to be read as hypothetical, preliminary, or as a statement of
biological fact
not needing experimental support. Extensive testimony was
presented on the
accepted scientific practices related to reporting data,
hypotheses, or
assumptions which are unsupported by experimentation. The
testimony of Drs.
Dustan, Hall, and Guyton all indicated that such data must
be clearly
identified as unsupported by experimentation. See Tr. Jan. 15,
1986, pp. 71
and 122; Tr. Jan. 14, 1986, pp. 111-114. It does not matter
that subsequent
experimentation may have confirmed that the Respondent's
published results
could have been obtained through experimentation. Tr. Jan.
14, 1986, pp. 174-
76. The Respondent tried to compare his presentation of
data with that of Dr.
Guyton, the chairman of the UMMC Department of
Physiology, in various
textbooks. I find that there is a clear distinction
between Dr. Guyton's
publications and the Respondent's publications. Dr.
Guyton's presentations
indicate even to the non-scientific reader that the
reported information is
generalized and does not refer to any specific
experimental results. See,
e.g., Exs. 100, 122; see also, Tr. Jan. 14, 1986,
pp. 207-11. The Respondent's
presentations, in contrast, appear to be
reporting actual results of
particular experiments.
2. I found
that the Respondent published inaccurate data and descriptions of
procedures
in the MMBG paper and the MHM paper, based on experiments in which
he had no
real personal involvement. Finding of Fact Nos. 7, 8, 9, 10, 11, 16
and 20.
While no evidence would support a finding that personal involvement in
reported experiments is directly required by any standard of conduct, the
general duty to report data and procedures accurately implies a need for
whatever degree of personal knowledge of data and procedures is necessary to
ensure accuracy. [FN16] The degree of personal knowledge necessary would
vary
depending upon the reliability and scope of the source of the
information
reported. When the source is highly reliable, such as a
published account, all
that may be necessary is to have read the published
account and to ensure that
it is accurately described. But when the source
is as unreliable as those used
by the Respondent, personal knowledge must be
secured by either personal
involvement, inquiries to those actually involved
in the experiment, or active
oversight and observation.
The
sources used by the Respondent were not sufficient in scope or
reliability
to provide the Respondent with sufficient personal knowledge of
the
experiments to ensure accuracy in the reporting of data and procedures.
For
the MMBG paper, even if I accept the Respondent's contention that he had
Dr.
Bengis' raw data and had performed tests on the blood samples in his own
laboratory, the Respondent's description of that data indicates the data was
not in a form that could be reliably reported without some greater knowledge
of the experimental procedures used. The Respondent stated that he
identified
the data values only by label markings from the test-tubes, and
that he
interpreted the data by making what were, in essence, educated
guesses
concerning which blood samples came from particular rats and at what
stage of
the experiment those samples had been taken. The Respondent never
asked Drs.
Bengis or Coleman about the precise meaning of the labels on the
samples, or
provided a copy of his manuscript to either to obtain comments
prior to
publication.
The Respondent admitted that he was not
certain of some important details of
the experiment, but did not consult
with any of the most obviously reliable
sources, such as Drs. Bengis and
Coleman or the other works reporting on these
same experiments. As I
discussed above, the Respondent stated that he did not
know the exact number
of subjects in each experimental group, but he did not
ask Drs. Bengis or
Coleman about the correct number. He based his published
account on the
impression of his laboratory technician, who had not been
directly involved
in the experiments. At the very least, when faced with
uncertainties in his
sources of information, the Respondent could have
consulted with Dr. Bengis'
thesis, or sought comments on the manuscript of his
paper. The Respondent
deprived himself of an important opportunity for input
by failing to consult
these other sources of information. In particular,
seeking comments on the
manuscript from other listed authors is a common
practice both to obtain
input and as a courtesy. Tr. Jan. 15, 1986, pp. 66 and
118.
3.
None of the other scientists who testified before me indicated that there
were any qualifications or conditions applicable to the general standard
that
experimental procedures and data must be accurately reported. None of
the
other scientists who testified before me had any confusion as to the
meaning
of the terms used by the Respondent, or the fact that the Respondent
inaccurately reported the experiments in the MMBG and MHM papers.
The Respondent persistently asserted that any inaccurate reporting was
either inadvertent or due to his use of an alternatively acceptable method
of
scientific reporting. His assertion that he reported only "scientific"
and
"biological" facts, and thus did not need to have supporting
experimental
data, was not supported by the testimony of any other
scientist. As I
discussed above, there are definite conventions governing
the reporting of
data unsupported by experimentation, whether it is
hypothetical, preliminary,
or otherwise. See, also, Tr. Jan. 14, 1986, pp.
174-76.
I can only conclude that the Respondent was unaware of or
simply disregarded
the generally accepted practices in the reporting of
research experiments.
B. STANDARDS OF CONDUCT FOR A PRINCIPAL INVESTIGATOR
4. THE RESPONDENT FAILED, AS A PRINCIPAL INVESTIGATOR, TO EXERCISE
ADEQUATE
OVERSIGHT OVER PROJECTS SUPPORTED BY GRANT FUNDS TO ENSURE THAT HE
REPORTED ON
THE RESEARCH CONSISTENT WITH ORDINARY STANDARDS OF CONDUCT FOR
REPORTING
SCIENTIFIC RESEARCH.
Discussion of the Standards of Conduct for a Principal Investigator
4. A principal investigator has a particular responsibility for
activities
under a research grant. The principal investigator is "a
qualified individual"
who is "responsible to grantee institution officials
for the proper management
and conduct of the project or program." Public
Health Service Grants Policy
Statement (rev.), October 1, 1976, p. 5. The
principal investigator "is
responsible for the scientific and technical
direction of the project." 42 CFR
52.2.
Testimony of Mary L.
Miers, an NIH employee, addressed more specifically the
responsibilities of
a principal investigator under a research grant. "[T]he
principal
investigator is responsible for supervising the work of others on
the
project, and in order to attain the status of principal investigator, is
supposed to know how to design a project, to interpret the data, [and] how
to
present them truthfully in publications or reports." Tr. Jan. 15, 1986,
p. 91.
This testimony was not rebutted or refuted. Although the witness
freely
admitted that she was not, herself, a scientist, her responsibilities
at NIH
and dealings with scientists on similar matters persuade me that she
could
accurately state a general understanding about the responsibilities of
a
principal investigator.
Even if I did not rely upon Ms. Miers'
testimony, I would have concluded
that these responsibilities were a natural
extension of the standards of
conduct for ordinary research scientists which
I discussed above, and were
implicit in the term "principal investigator."
In the grant application, the
Respondent signed a statement that, as
principal investigator, he would
"accept responsibility for the scientific
and technical conduct of the
research project...." See Ex. 227, p. 1. This
statement indicates acceptance
of more than just the responsibility for the
principal investigator's own
work, but also implies that the principal
investigator must oversee or
otherwise ensure the accurate reporting of any
work actually carried out by
others with the support of federal grant
funds.
The Respondent argued that he had reasonably relied on others
for
information to ensure accurate reporting of the experiments in question,
including information on the results of the experiments and the procedures
followed. Considering the lack of any actual involvement in the experiments,
and the lack of any demonstrated oversight or supervision of the conduct of
the experiments, I find that reliance to have been somewhat unreasonable
when
the Respondent did not seek to obtain definitive, updated information
from
sources directly involved in the experiments. In essense, the
Respondent
attempted to absolve himself of blame by placing the
responsibility on others
to keep him informed and to correct his
mistakes.
With respect to the MMBG paper, for example, the Respondent
was not on Dr.
Bengis' dissertation committee and did not exercise any of
the supervisory
oversight which the role of a principal investigator
logically requires. The
Respondent stated that, "[m]y knowledge of Dr.
Bengis' research study was
obtained by visiting his laboratory during the
time he was performing his
studies." Ex. 146, p. 3. But, the Respondent's
lack of direct involvement in
the experiments, apart from these visits,
would make it possible that the
experimental protocol was altered without
his knowledge. The Respondent
testified that he relied merely on information
gleaned from test-tube labels
and raw data sheets which did not, apparently,
have the degree of reliability
necessary to ensure accuracy in interpreting
data values.
The same type of unreasonable reliance was present in
connection with the
MHM paper. The Respondent testified that the research
was conducted
independently and that his only source of information was a
vague graphic
representation for which the Respondent stated that he was not
responsible.
Even if I accepted the Respondent's own account of what he
relied on, I would
conclude that he failed to exercise the oversight and
caution which the role
of principal investigator requires.
The
Respondent asserted that he was not responsible for inaccurate reporting
which occurred because of failures of others to correct the galley drafts or
failures of others to inform him of changes in procedures. The testimony
presented concerning which author of an article was responsible for making
corrections was inconclusive. See, generally, Ex. 160, pp. 24-30. Although
standards of responsible conduct appear to be evolving toward placing
greater
burdens on the first author of an article, the significance of being
listed as
first author and the standards for the role of that author are not
uniform.
But, the Respondent was not merely a collaborating scientist on
these papers;
he was presenting the work as a part of the research project,
supported by
federal grant funds, for which he was principal investigator.
In an ordinary
circumstance, he may not have been required to assume a
larger responsibility
than other authors. As principal investigtor under a
research grant, however,
he had a greater burden than his colleagues to
ensure the veracity and
accuracy of anything he published.
I do
not mean to suggest that a research scientist can never rely upon
colleagues, students, and employees. When a research scientist relies upon
others, however, that reliance must be reasonable under the circumstances.
The
reliance must be based on a working relationship with sufficiently good
communication to ensure accurate reporting of the final results.
Furthermore,
the principal investigator bears an additional burden to ensure
that the
project was carried out as planned and that the results to be
reported were
actually obtained.
The Respondent did not have
reasonable reliance in these circumstances. Even
accepting his own
descriptions of his documentary sources of information at
the time he wrote
the MMBG and MHM papers, these sources were not sufficiently
detailed that
the Respondent could reasonably have believed that he could
accurately
report the experiments in question, without supplemental
information from
those who conducted the experiments. The record clearly
indicates that the
Respondent did not maintain lines of communication to
obtain any further
input. Although he stated that he was frequently in Dr.
Bengis' laboratory,
his lack of knowledge about key details of the experiments
show that he had
little actual involvement in either planning the Bengis
experiments or in
the conduct of the experiments. Similarly, he was not
involved at all in the
conduct of the Hall experiments. Although his working
relationship with Dr.
Bengis is a little uncertain, the Respondent's own
testimony was that his
working relationship with Dr. Hall was strained.
Furthermore, the
Respondent knew or should have known that the results he
reported were not
consistent with the actual experiments. The results of the
Bengis
experiments were available to the Respondent in Dr. Bengis' thesis, and
the
Respondent admitted that Dr. Hall had informed the Respondent that a
relevant portion of the MHM paper was inaccurate. The Respondent was
responsible, as principal investigator, to secure sufficiently reliable
support for his reporting. Obviously, he did not do so, even when it should
have been apparent to him that the information he says that he had would not
be sufficient.
C. REASONS FOR DEVIATIONS FROM STANDARDS OF CONDUCT
5. THE RESPONDENT'S DEVIATIONS FROM THE STANDARDS OF CONDUCT DESCRIBED
ABOVE, FOR BOTH ORDINARY RESEARCH SCIENTISTS AND PRINCIPAL INVESTIGATORS,
WERE
CAUSED BY A NEGLIGENT LACK OF DUE CARE AND, WITH RESPECT TO CERTAIN
DEVIATIONS, AN INTENT TO DECEIVE.
Discussion of the Reasons for Deviations from Standards of Conduct
5. With respect to the inaccurate presentation of procedures and
results in
the MMBG paper, I found that the Respondent knew or should have
known that he
did not have sufficient knowledge of the experimental
procedures and results
to accurately report the methods and results set
forth in the paper. Finding
of Fact No. 13. With respect to the inaccurate
presentation of the number of
subjects, certain values, and the statistical
analysis in the MHM paper, I
found that the Respondent knew or should have
known that he had inaccurately
reported these elements in the MHM paper.
Finding of Fact No. 24.
In light of my finding that the standards of
conduct exercised by ordinary
research scientists require accurate reporting
of experimental procedures and
data, I must conclude that the Respondent was
negligent in not adhering to
these standards.
With respect to the
presentation of certain results and descriptions of
procedures in the MMBG
paper, the MHM paper, the subsequent works, and the
grant applications, I
found also an intent to mislead the reader into a belief
that the Respondent
had performed the reported experiments and obtained the
reported results.
Finding of Fact Nos. 14, 25, and 32.
The Respondent apparently
intended to deceive readers in order to enhance
his reputation in the
scientific community and to better his prospects for
receiving federal grant
funds. He sought to give the impression that he was
more active in his field
than he actually was, and had more involvement in
experimental work than he
actually had.
D. UNSATISFACTORY PERFORMANCE OF GRANTS
6. THE RESPONDENT'S FAILURE TO ADHERE TO THESE STANDARDS OF CONDUCT,
AND HIS
FAILURE TO HAVE SUFFICIENT KNOWLEDGE ABOUT THE REPORTED RESEARCH TO
ENSURE
THAT HE ADHERED TO THESE STANDARDS OF CONDUCT, RESULTED IN
UNSATISFACTORY
PERFORMANCE UNDER THE RESPONDENT'S GRANT AWARDS.
Discussion of the Unsatisfactory Performance of Grants
6. In the applications for research grant HL-09921, the Respondent
established the long-term goals of the research program to be supported by
the
grant. Those goals were to broadly study all phases of aldosterone
secretion,
distribution, and metabolism, and to study aldosterone's
influence on sodium
and fluid reabsorption. Ex. 261, p. 7; Ex. 262, p. 6;
Ex. 263, p. 2; Ex. 265,
p. 3; Ex. 266, p. 8.
Implicit in the grant
proposals were the assumptions that the work would
conform to the ordinary
standards for such research programs and would be
published or otherwise
reported to the scientific community. I found no
references in the grant
applications to indicate that the Respondent was not
offering to perform
publishable research conforming to ordinary standards.
While there may be
circumstances when a grant may be given for preliminary,
secret, or other
unpublishable research which may not be held to ordinary
research standards,
there was no indication in the applications that the
Agency was agreeing to
any special standards. Indeed, there are numerous
references in the
applications to the publications record of the Respondent,
and to the impact
of the research in the scientific community as a whole. See,
e.g., Ex.
227.
I have found above that the Respondent did not conduct his
research program
in accordance with the standards of conduct employed by
ordinary research
scientists. Conclusion of Law Nos. 1, 2, 3, and 4. The
Respondent's failure to
conform to these standards of conduct was not beyond
his control to avoid.
Even though my findings above are based on only
some of the Respondent's
published works, the problems identified in those
works raise questions about
the reliability of all of the Respondent's
works. This is particularly so
because the Respondent, in explaining the
identified errors, made clear his
basic disregard for accurate reporting of
experimental research.
Since the Respondent has thus called his entire
body of work into question,
all of which was supported by his research
grant, I conclude that the work was
not a satisfactory performance of the
Respondent's duties under the grant.
D. LACK OF RESPONSIBILITY USING GRANT FUNDS
7. THE RESPONDENT'S FAILURE TO ADHERE TO THESE STANDARDS OF CONDUCT,
AND HIS
FAILURE TO HAVE SUFFICIENT KNOWLEDGE ABOUT THE REPORTED RESEARCH TO
ENSURE
ADHERENCE TO THESE STANDARDS OF CONDUCT, INDICATES A LACK OF
RESPONSIBILITY.
Discussion of the Lack of Responsibility Using Grant Funds
7. The Respondent's actions formed a pattern of deception which I
consider
in assessing the Respondent's present responsibility to
satisfactorily perform
under a federal research grant. In general,
responsibility requires
compentence, capacity, a record of satisfactory
performance, integrity, and
ethics. See, generally, 48 CFR 9.104-1 et
seq.
In evaluating the Respondent's capacity and competence, I have
already
concluded that the Respondent's past performance record was
unsatisfactory,
but I must also consider the reasons for that record. See
Conclusion of Law
No. 6; Marine Engineer's Beneficial Association, 39
Comp.Gen. 705 (1960).
Neither party appears to contend that the Respondent
lacked the ability to
perform satisfactorily. The Respondent is an
intelligent and experienced
scientist who several of the Agency's witnesses
freely admitted had made
important contributions to his field. E.g. Tr. Jan.
14, 1986 pp. 43-45; Ex.
160.
The Respondent, however, lacked the
tenacity and perseverance to exercise
the amount of care necessary to ensure
accurate reporting of experiments. For
example, the Respondent failed to
exercise the personal discipline and care
necessary to check with his
original sources when uncertain about particular
facts and details in both
the MMBG and MHM papers. See Finding of Fact Nos.
13, 14, 21, 22, 23, 24 and
25. Even when informed of an error in the MHM
paper, he failed to follow
through to ensure that the error was corrected, and
failed, in subsequent
publications of the same experimental data, to correct
the error.
In evaluating the Respondent's integrity and ethics, I have already
concluded that the Respondent violated standards of ethical conduct for
research scientists. Conclusion of Law No. 1. With respect to some of those
violations, the Respondent had an intent to deceive and was not merely
negligent. Conclusion of Law No. 5. In particular, I note that the
Respondent
knowingly submitted grant applications with information which he
misrepresented as the results of actual experiments when those experiments
were not performed either by him or by any of his collaborators. While I do
not presume to make any conclusion of culpability under laws, such as 18 USC
1001, which would be more appropriate for a court, I find that the record
indicates a pattern of misrepresentation which is inconsistent with any
presumption of honesty and integrity.
F. CAUSES FOR DEBARMENT ESTABLISHED
8. RESPONDENT'S FAILURE TO ABIDE BY HIS OBLIGATIONS AS A PRINCIPAL
INVESTIGATOR VIOLATED THE TERMS AND CONDITIONS OF THE AWARD OF GRANT FUNDS
AND
ESTABLISHES A CAUSE FOR DEBARMENT UNDER 45 CFR 76.10(d).
9.
THE RESPONDENT'S RECORD OF UNSATISFACTORY PERFORMANCE ESTABLISHES A CAUSE
FOR DEBARMENT UNDER 45 CFR 76.10(e).
10. THE RESPONDENT'S PATTERN
OF DISREGARD FOR CERTAIN STANDARDS OF CONDUCT
OBSERVED BY RESEARCH
SCIENTISTS IS A "CAUSE SIGNIFICANTLY AFFECTING
RESPONSIBILITY AS A RECIPIENT
... UNDER A FEDERAL PROGRAM OF SUFFICIENTLY
SERIOUS NATURE ... TO WARRANT
DEBARMENT" UNDER 45 CFR 76.10(g).
Discussion of the Establishment of the Causes for Debarment
8. The research grant under which the Respondent received funds had
few
formal requirements of relevance here. In general, this type of research
grant
depends heavily upon the trust which the government places in the
principal
investigator, to direct and oversee the research conducted with
the support of
grant funds. The proposed debarment is based, in part, on the
failure of the
Respondent to fulfill his duties as a principal investigator
with regard to
research supported by the grant.
I have concluded
above that the Respondent failed, as principal
investigator, to ensure that
he reported the research projects supported by
federal grant funds in
accordance with the ordinary standards of conduct for
reporting scientific
research. Conclusion of Law No. 4. The Public Health
Service Grants Policy
Statement (rev.) October 1, 1976, makes clear that
reporting on work
performed under a grant is an activity within the terms and
conditions of
the grant. Annual progress reports must be made to the
government, and
"[p]roject directors and principal investigators are
encouraged to make the
results and accomplishments of their activities
available to the public."
PHS Grants Policy Statement, pp. 41, 74.
The Respondent's failure to
fulfill his responsibility as a principal
investigator was, therefore, a
violation of the terms and conditions of his
grant award. It is particularly
egregious in view of the large degree of
responsibility and trust placed in
the principal investigator under the terms
of the grant.
The
Respondent's serious violations of the terms and conditions of the grant
award establish a cause for debarment under 45 CFR 76.10(d), which permits
debarment in the public interest for: "[s]erious violation of the applicable
statute, regulations or other terms and conditions of a previous award of
financial assistance."
9. I have concluded above that the
Respondent's failure to adhere to certain
standards of ordinary research
conduct, and his failure to have sufficient
knowledge about reported
research to ensure that he adhered to these
standards, resulted in
unsatisfactory performance under the Respondent's grant
awards. Conclusion
of Law No. 6.
The instances of such unsatisfactory performance were
not isolated and
discrete, but formed a continuous pattern over a period of
years which was
central to the research program which the Respondent
asserted that he was
pursuing in his grant applications and publications.
The inaccurate reporting
in the MMBG and MHM papers was repeated in grant
applications and, for the MHM
paper, in subsequent publications and
manuscripts. These repeated actions
establish a record of unsatisfactory
performance, within the Respondent's
control to avoid.
The
Respondent's record of unsatisfactory performance establishes a cause
for
debarment under 45 CFR 76.10(e), which permits debarment in the public
interest for: "[a] record of serious unsatisfactory performance (or failure
to
perform) under one or more prior awards of financial assistance
..."
10. I have concluded that the Respondent did not act responsibly
in failing
to adhere to certain standards of research conduct and in failing
to have
sufficient knowledge to ensure adherence to those standards for
research under
his grant. Conclusion of Law No. 7.
Therefore, I
conclude that the pattern of disregard for ordinary standards
of research
conduct and duties under grant terms is a cause for debarment
because it is
behavior "significantly affecting responsibility as a recipient
... under a
federal program of sufficiently serious nature." 45 CFR 76.10(g).
G. NO SUBSTANTIAL MITIGATING FACTORS OR CHANGE IN RESPONDENT
11.
THERE EXIST NO MITIGATING FACTORS TO SUPPORT A DETERMINATION THAT
DEBARMENT
IS UNWARRANTED.
12. THE RECORD CONTAINS NO PERSUASIVE EVIDENCE OF ANY
CHANGE IN THE
RESPONDENT'S CONDUCT, ABILITY, OR ATTITUDE WHICH WOULD
INDICATE THAT THE
RESPONDENT IS PRESENTLY RESPONSIBLE AND CAPABLE OF
SATISFACTORY PERFORMANCE.
Discussion of Mitigating factors and Changes in the Respondent
11. The regulations state that I must consider whether debarment is
warranted in light of "mitigating factors." 45 CFR 76.11. The Respondent did
not argue that there were any mitigating factors, but I have found elements
in
the record which I will consider in determining whether there are
mitigating
factors which should affect a decision to debar. In my
deliberations, I have
considered the consequences of both a broad and a
narrow interpretation of the
term "mitigating factors," and have found no
evidence which persuades me that
debarment is unwarranted.
Considering factors which could be considered "mitigating factors," I find
no cause to refrain from debarring the Respondent. I have found that some
evidence in the record suggests, generally, that the Respondent's actions
may
have been affected by career related pressures, problems in the
Respondent's
personal life, and problems related to changes in the physical
location of the
Respondent's laboratory facilities. Finding of Fact No. 3.
No evidence
suggests that these circumstances are unusual, unforeseeable, or
unlikely to
occur in the life of a research scientist. Thus, I do not find
that these
circumstances can, in any sense, constitute an excuse for the
Respondent's
actions, or even satisfactorily explain those actions. It
appears from the
testimony before us about the standards of scientific
conduct that other
research scientists subject to similar pressures are
capable of maintaining
the integrity of their work.
I find no
mitigating factors in the Respondent's lengthy presentations
concerning the
atmosphere of scientific impropriety which he alleged was
created by the
wrongdoings of others at UMMC. The supposed wrongdoing of
others has no
direct bearing on the relationship between the Respondent and
the government
agency funding the research.
Considering only the degree and
seriousness of the Respondent's wrongful
acts, I find no cause to refrain
from debarring the Respondent. I find above
that the Respondent's acts were
not merely technical violations of regulatory
requirements, but were instead
violations of the most fundamental standards of
conduct. These acts
indicated a disregard for the fundamental purpose of
research and a
disregard for the responsibilities implicit in the position of
principal
investigator. See Conclusion of Law Nos. 1, 2, 3, 4 and 5.
In
extending grant funding for a program of research into cardiovascular
systems, Congress signified the national importance of such research. See 42
USC 241 and 287a. Congress mandated that the NHLBI should foster and
coordinate a research program under which research findings would be
published
and disseminated in a timely manner: both the raw findings and any
practical
applications of those findings. 42 USC 287a(5) The requirement
that research
results be published and disseminated makes it clear that the
government and
others will rely on the integrity of that research.
While the record does not indicate a specific instance of any such reliance,
I do not find that a specific instance is necessary for the Respondent's
lack
of research integrity to constitute a serious offense. Although there
is an
obvious possibility that the government relied on the Respondent's
work in
support of an application for the marketing of captopril, the Agency
did not
argue that this had occurred. But it is sufficient that the
government may
have relied upon the integrity and accuracy of the
Respondent's works to
determine the course of its research program,
including any decision to extend
further funding to the Respondent. See 42
CFR Part 52.
It does not matter, in this context, whether the reported
results were what
would have been obtained by conducting the experiments.
That would not change
the underlying failure to adhere to standards of
conduct which would be
followed by ordinary research scientists, lack of
knowledge of generally
accepted research practices, failure to exercise
adequate supervisory
oversight as principal investigator, and intentional
attempts to deceive the
government and the larger scientific community to
obtain greater recognition
and continued grant funding.
12. The
Respondent has consistently denied any wrongdoing and, in these
denials, he
has indicated that his perception of the responsibilities and
duties of a
scientific researcher receiving federal grant funds has not
changed. I have
found above that the Respondent admitted to the basic facts
underlying the
inaccuracies in the two papers which are at the core of this
inquiry--the
MMBG paper and the MHM paper. The Respondent's own description of
the manner
in which he wrote these papers shows his lack of care and
inattention to
detail in the failure to check sources and ensure accuracy.
Although
the Respondent made some attempts to retract those papers which had
been
called into question, he did so only at the request of others. His
consistent position has been that, even though the papers may have been
unsupported by actual experimental data, there was no need to retract the
papers since the results reported would have been obtained if the
experiments
had been performed. The Respondent claimed that the experiments
were
unnecessary since the results were "scientific" or "biological" facts.
However, the Respondent was unable to point to any specific, published
support
that existed at the time which had established these facts, and did
not, in
fact, cite any works which established these facts in his
papers.
The Respondent's apparent belief that there are exceptions to
this basic
scientific standard indicates that the Respondent's present
attitude and
ability have not changed with respect to conducting and
reporting scientific
research. As I stated above, in the course of the
hearing the testimony of
scientists established that the most basic standard
of scientific conduct is
that experimental procedures and results must be
accurately reported to permit
critical evaluation and replication of the
experiment. Even if a scientist
intuitively believes that a result is
inevitable, that result may only be
reported as an experimental result if
the experiment was actually performed.
See Conclusion of Law No. 1. Only the
Respondent has suggested that this
standard should not apply when it can be
subsequently established that the
reported result was correct.
Dr.
Guyton asserted his belief that the Respondent had acted honestly at
least
since 1982, when the doubts which led to this proceeding arose. Tr. Jan.
14,
1986, pp. 116-20. Dr. Guyton also stated his belief that the Respondent's
work would be reliable in the future, because the Respondent has learned
"his
lesson." Id., p. 120. While I sincerely hope that this is true, nothing
in the
record provides support for Dr. Guyton's belief. Even Dr. Guyton was
unable to
provide any demonstrable evidence of a change in the Respondent.
Id., p. 118.
H. DEBARMENT IS WARRANTED BY THE CIRCUMSTANCES OF THIS CASE
13. THE
RESPONDENT IS NOT PRESENTLY RESPONSIBLE AND THE INTERESTS OF THE
FEDERAL
GOVERNMENT REQUIRE THAT HE BE DEBARRED FROM RECEIVING AWARDS OF GRANT
FUNDS.
14. DEBARMENT FOR THE FULL TERM OF THREE YEARS IS WARRANTED
BY THE
SERIOUSNESS OF THE ACTS INVOLVED AND IS NECESSARY TO PROTECT THE
INTERESTS OF
THE GOVERNMENT.
Discussion of Whether Debarment is Warranted
13. I have concluded above that the Respondent's actions showed a lack
of
responsibility, and that there is no persuasive evidence that the
Respondent's
conduct, ability, or attitude have changed in a way that would
indicate that
the Respondent was presently responsible. Conclusion of Law
Nos. 7 and 11.
In light of the high degree of trust inherent in
research grants of the type
the Respondent has received, under which
performance cannot be readily
verified or qualitatively monitored, I
conclude that the interests of the
federal government require that he be
debarred. There is no lesser sanction
that would adequately protect against
further intentional inaccurate reporting
of experimental procedures and
results.
I note, additionally, that the harm in the Respondent's
actions does not
merely stem from a waste of government money for research
activities. Just as
the Respondent's actions have called into question the
veracity of his entire
range of works, so the presence of a few inaccurate
papers causes irreparable
harm to the credibility of scientific research
papers in general, and the
entire government research effort. As Dr. Roscoe
testified, the scientific
community functions on trust and openness, and
once that trust is breached,
the harm is far greater than the loss of the
accurate presentation of research
in one or two papers.
14. The
three-year term proposed for the debarment is a standard length for
an
administrative debarment. See, e.g., Policy Letter No. 82-1, 47 Fed.Reg.
28854 (June 24, 1982). I have the discretion, however, under the regulations
to select any "reasonable, specified period of time, commensurate with the
seriousness of the cause resulting in the debarment." 45 CFR 76.2.
[FN17]
In this case, the offenses were of a highly serious nature and
I have found
no mitigating factors which would affect a decision to debar or
the term of
the debarment. The offenses related directly to the Respondent's
integrity,
his ability to conform to accepted scientific practices, and his
performance
in accordance with the terms and conditions of the grants
awards.
The offenses took place over a period of time: the reported
research
projects were conducted during 1976-79; the Respondent published
the papers
giving rise to this action in 1978 and 1979, and reiterated the
inaccurately
reported material in subsequent publications, manuscripts, and
grant
applications as late as September 1981.
I conclude that the
three-year proposed debarment period is reasonable and
commensurate with the
seriousness of the causes for debarment outlined above.
In determining from
what date to measure the commencement of this debarment
period, I have
considered a number of factors, including that the Respondent's
grant was
terminated on May 10, 1985, that the Respondent was initially
notified of
the debarment proceeding on August 15, 1985, and that the process
beginning
with the anonymous allegations in 1982 and leading to this debarment
has
been lengthy, in part, because it has been the first contested debarment
of
a research scientist by this Department.
Nevertheless, I reject the
Respondent's argument that any debarment should
date from May 10, 1985.
Respondent's Post-Hearing Brief, p. 98. Although that
was the last date that
the Respondent received grant funds, this termination
date was not required
by the federal government, but was an independent action
by UMMC. Exs. 43
and 149. There was no "effective debarment" or suspension.
From the federal
government's perspective, the Respondent's funding could have
continued
through June 30, 1985. Beyond that date, the Respondent had not been
debarred or suspended either. Although NIH had determined not to renew the
grant, with Respondent as principal investigator, for additional budget
periods, he could have applied for other grant funds and NIH would have been
required to consider such an application. Ex. 149.
Although I have
rejected the Respondent's arguments regarding the
commencement date for the
debarment, I have determined not to set the
commencement date at the date
this decision is issued, but, instead, at the
date this decision would
likely have been issued if not for some unforeseen
delays in scheduling the
in-person hearing and in transcript preparation. The
Agency agreed that
there was a reasonable basis for a 5-6 month reduction to
account for these
delays. Agency's Post-Hearing Brief, pp. 56-57.
Accordingly, in order
to strike a proper balance between protecting the
Government's interests and
not unduly disadvantaging the Respondent by these
unforeseen delays, I have
determined to set the commencement date for the
Respondent's three-year
debarment at July 15, 1986.
VII. Conclusion
For the reasons stated above, I have determined that the Respondent
should
be debarred for a period of three years beginning July 15, 1986.
Cecilia Sparks Ford
Hearing Officer
Member, Departmental Grant Appeals
Board
FN1 These terms were used in their ordinary sense by the Debarring Official
in
the proposal to debar. In the body of the decision, in general, I have
used
terms which are narrower in scope and more precise.
FN2 As noted above, the contract debarment regulations provide for a fact-
finding proceeding when there is a "genuine dispute over facts material to
the
proposed debarment." 48 CFR 9.406-3. Early in this proceeding, Agency
counsel
argued that the Respondent was not disputing the "underlying"
material facts
relating to the proposed debarment. The Agency characterized
the Respondent's
presentation as an attempt to explain or to shift the blame
to others with
regard to the incidents of scientific misconduct found by the
NHLBI panel
which investigated this matter. Agency's Pre-Hearing Brief, pp.
16-20. The
Respondent had asserted, however, that "[t]he NIH Report is
neither accurate
nor factual" and that this report misrepresented the facts.
Exs. 73 and 88. I
viewed these assertions together with Respondent's general
denial that he
"fabricated research data or falsified drug studies" as
raising a genuine
dispute over material facts. Ex. 73.
FN3 The four members of the NHLBI panel were scientists with distinguished
credentials and experience in the field of hypertension and related
functions.
Tr. Jan. 15, 1986, p. 40. Dr. Dustan, the panel chair, was the
director of the
Cardiovascular Research and Training Center at the
University of Alabama
Medical Center. Dr. Gomez-Sanchez was a professor of
Medicine at the
University of South Florida School of Medicine. Dr. Knox was
the Dean of the
Mayo Medical School and Director for Education of the Mayo
Foundation. Dr.
Mulrow was the Chairman of the Department of Medicine at the
Medical College
of Ohio. See Agency Hearing Exs. D-G (Curriculum Vitae).
FN4 The UMMC subsequently terminated the Respondent's employment for cause,
based upon the NHLBI report and its findings. A hearing was held prior to
the
termination. The conduct of that hearing and the propriety of the
Respondent's
termination are not at issue in the case before me.
FN5 The record contains a report of a polygraph examination of Repondent. Ex.
95. This report was submitted to NIH on October 3, 1984 as evidence of the
Respondent's truthfulness "concerning the allegations brought against him."
I
do not consider the report useful in this matter. Polygraph examinations
in
general are of doubtful evidentiary worth. See, e.g., Barrel of Fun, Inc.
v.
State Farm Fire Casualty Co., 739 F.2d 1028 (5th Cir.1984); United States
v.
Alexander, 526 F.2d 161 (8th Cir.1975). Moreover, the questions set forth
in
the report as the subject of the examination were not particularly
illuminating.
FN6 The scientific misconduct of others was not at issue in this proceeding
and is subject to separate NIH review. I admitted evidence into the record
concerning the alleged wrongdoing of others for the limited purposes of
assessing the credibility of the witnesses and of establishing standards
within the scientific community.
FN7 The Respondent did not cite the final paper written by Drs. Bengis and
Coleman; that paper was published after the MMBG paper.
FN8 Dr. Bengis asserted that, generally, he gave no samples to the
Respondent's laboratory for testing. He admitted, however, that he gave at
least some preliminary samples to the Respondent's laboratory for renin
testing, to compare results from the Respondent's laboratory with results
from
Dr. Cowley's laboratory. Tr. Jan. 15, 1986, p. 22. But these samples,
he
stated, were all samples taken from dogs. Tr. Apr. 3, 1986, pp.
19-20.
FN9 While Dr. Bengis testified that he may have had occasion to bring samples
from other experiments to Dr. McCaa's laboratory for aldosterone assay, this
testimony alone was not sufficient to rebut the more detailed testimony of
Respondent's witnesses. Tr. Apr. 3, 1986, pp. 38 and 45.
FN10 In reviewing this testimony, I find it to be inconclusive and relevant
only to show that some witnesses appeared more willing to admit to the kind
of
gaps in memory which I would consider normal when recalling events nine
years
past.
FN11 Each of the experimental groups had a set of control measurements, taken
before captopril administration from subjects treated identically as other
experimental group members. The 'control group' was a separate group of rats
not treated with dietary or surgical modifications.
FN12 The Respondent further explained that he had obtained the number 12 by
checking with Ms. Gray, his former laboratory technician, because he had not
been certain of the correct number. Ms. Gray's testimony indicates that she
never had a substantive knowledge of the procedures used in the experiment
and
was involved only in the mechanics of the aldosterone testing. She
testified
that she did not analyze the raw data, identify that data with any
specific
experimental groups or procedures, or even record the data in final
form. She
merely ensured that the data was properly identified with the
proper test
sample. Given Ms. Gray's lack of knowledge concerning
experimental procedures,
I can not find that the Respondent had a reasonable
basis to believe that Ms.
Gray was a reliable source of information for the
number of animals in each
experimental group, particularly since better
sources of that information were
available.
FN13 I have excluded from my basis for this finding several other reported
data points which differed from the data reported by Drs. Bengis and Coleman
but were not sufficiently different for it to be impossible that they came
from the same experiments.
FN14 The record contains numerous allusions to interpersonal conflicts which
may have caused the Respondent and Drs. Bengis and Coleman to be less than
open with each other. This might explain why the Respondent may have
conducted
aldosterone assays even though the protocol no longer required
them, and why
the Respondent, apparently, never reported aldosterone results
to Drs. Bengis
and Coleman.
FN15 Dr. Horowitz is Vice President of the Squibb Institute for Medical
Research. Squibb provided the Respondent with SQ-14,225 for experiments at
UMMC. Dr. McCaa corresponded with Dr. Horowitz to report research results
and
obtain drug samples. Tr. Apr. 3, 1986, pp. 50-51.
FN16 No evidence or argument was directed at the standards of conduct for
attribution or citation of scientific research. It may be that a certain
degree of personal involvement in the experiment is necessary to report
research data and procedures without incurring a duty to acknowledge that
the
work was performed principally by others.
FN17 Although only tangentially relevant to this case, I note that the
debarment periods in two voluntary settlements with research scientists (the
only other such debarments discussed in this record) were longer, being for
ten-year and four-year terms. Tr. Jan. 15, 1986, p. 80.