What Is Ethically Important in Clinical
Research? A Preliminary Study of Attitudes of
73 Psychiatric Faculty and Residents
by Laura Weiss Roberts, Teddy D. Warner, Janet L. Brody, Khanh Nguyen,
and Brian B. Roberts
damental to ethical human research: respect for persons,
beneficence, and justice (National Commmission for the
Protection of Human Subjects of Biomedical and Behavioral Research 1979). Respect for persons is the recognition of the research volunteer as a full, individual person
with a distinct life history, strengths, and values who
deserves human regard and dignity. Beneficence is the
professional duty to bring about good, or benefit, for individuals and for society through the application of specialized skill or expertise. Justice is the notion that benefits
and burdens should be equitably distributed in a shared
society or community; in research, this principle suggests
both the importance of providing access to positive
opportunities represented in research participation and the
obligation not to exploit certain groups or volunteers by
virtue of convenience or vulnerability. Key ethical justifications for specific research protocols include the potential benefit for severely ill participants; adequacy of safeguards such as informed consent, confidentiality
protections, and institutional review board (IRB) review;
and positive qualities of investigators (e.g., professional
integrity) and of the investigator-participant relationship
(e.g., trust) (Levine 1988; Department of Health and
Human Services 1991; Dresser 1996; Human Radiation
Experiments 1996; National Bioethics Advisory Commission 1998; Roberts 1999).
The ethical caliber of mental illness research ultimately rests with the individual professionals who perform this work and are responsible for safeguarding the
welfare of mentally ill individuals. Although psychiatrists
shoulder this responsibility, little is known about their ethical attitudes toward human experimentation. The aim of
this article is to present our findings on psychiatrists' attitudes toward 12 ethically relevant elements in the conduct
Abstract
The ethical caliber of psychiatric research ultimately
rests upon the shoulders of psychiatric clinicians and
investigators who perform protocols and are directly
responsible for the welfare of study volunteers. Little
is known, however, about ethically relevant attitudes
of psychiatrists toward human research. Working as
part of a larger study, we surveyed all psychiatry faculty and residents at one institution in 1999 regarding
the relative importance of 12 elements in the ethical
conduct of human research. Responses of 73 participants were analyzed, and three factors emerged:
Safeguards and Scientific Merit, Investigator Integrity
and Relationship, and Patient Benefit Safeguards and
Scientific Merit, as a set, were the most salient considerations for both psychiatry faculty and residents. We
found that residents placed greater importance on all
factors and nearly every element than did faculty.
Future research is needed to clarify the understanding
of the perspectives and priorities of different stakeholders involved in human studies.
Keywords: Clinical research, ethics, psychiatry,
scientific integrity, schizophrenia.
Schizophrenia Bulletin, 29(3):607-613,2003.
Ensuring the ethical conduct of mental illness research is a
commitment of utmost importance in society and the scientific community (Hoagwood et al. 1996; Hyman 1999;
Roberts 1999; Roberts 2002). Policy makers, scientific
leaders, and advocates have conscientiously worked
toward a shared understanding of acceptable standards for
human investigation, particularly in the context of studies
involving special populations (Dresser 1996; The Human
Radiation Experiments: Final Report of the President's
Advisory Committee 1996; Brody 1998; Rappaport and
Carolla 1999).
Three principles have come to be recognized as fun-
Send reprint requests to Laura Roberts, M.D., Professor and Chair,
Department of Psychiatry and Behavioral Medicine, Medical College of
Wisconsin, 8701 Watertown Plank Rd., Milwaukee, WI 53226; e-mail:
[email protected].
607
L.W. Roberts et al.
Schizophrenia Bulletin, Vol. 29, No. 3, 2003
of clinical research. These data derive from a larger project
involving 73 residents and faculty in psychiatry in which
we discovered that these psychiatrists strongly affirmed
the importance of schizophrenia research and identified
several positive motivations for participation (e.g., altruism, helping science, conferring hope) (Roberts et al.
2000). To our knowledge, no other data have been published regarding psychiatrists' perspectives on the ethics of
mental illness research.
main effect for respondent (F(l,67) = 5.8, p = 0.02)
revealed that residents (mean = 4.3) rated these elements
overall as more important than did faculty (mean = 4.0; d
= 0.26). Gender and other variables, such as ethics experiences, showed no reliable effects.
Dimensions of ethical elements (figures 1-3).
Exploratory factor analysis was performed to determine
whether the 12 elements were indicators of latent dimensions (Nunnally and Bernstein 1994). A varimax-rotated
three-factor solution (variance explained = 45%) appeared
to best fit the data and provide factors with optimal meaning. The first factor consisted of four items pertaining to
Safeguards and Scientific Merit (Cronbach's a = 0.68)
(figure 1). The second factor included five items pertaining to Investigator Integrity and Relationship (a = 0.65)
(figure 2). The third factor consisted of three items pertaining to perceived Patient Benefit (a = 0.77) (figure 3).
Composite variables were constructed based on the average of items loading on a factor.
A MANOVA, with these three factors as a within-subjects factor and respondent and gender as the between-subjects factors, was conducted. A main effect for factor
(F(2,66) = 89.8, p < 0.0001) was found. Safeguards and
Scientific Merit was rated as highly important (mean =
4.7), Investigator Integrity and Relationship was rated as
important (mean = 4.3; p < 0.01; d = 0.33), and Patient
Benefit was rated neutrally (mean = 3.2; p < 0.001; d =
1.25 and 0.93, respectively). A main effect for respondent
(F(l,67) = 7.2, p < 0.01) was detected. Faculty members
rated the three factors, overall (mean = 3.9) as less important than residents did (mean = 4.2; d = 0.24). Both main
effects are qualified by a significant interaction (F(2,66) =
3.3, p < 0.05) between respondent and factor. Residents
rated the first two factors as only moderately more important than did faculty (d = 0.26 and 0.45, respectively) but
rated the third factor, Patient Benefit, much higher than did
faculty (d = 0.80). Other potential moderating variables
showed no reliable effects.
Methods
In this IRB-approved study, all faculty and resident psychiatrists at the University of New Mexico School of Medicine
were asked in 1999 to complete a confidential survey that
assessed their attitudes toward human research. Respondents were not compensated. More information regarding
study methodology and other results from the overall study
have been reported elsewhere (Roberts et al. 2000, in press
a, in press b; Warner et al. 2002). For this report, participants were asked to indicate the importance (scale: 1 = not
at all important, to 5 = very important) of 12 elements in the
ethical conduct of clinical research (table 1).
Results
Characteristics and Ethics Experiences of Respondents. Overall, 83 percent of psychiatrists (n = 71; 46 faculty and 25 residents; 39 men, 28 women, 2 unspecified;
44% married; 75% white) provided responses to all items.
Participants reported spending a majority of their time in
clinical activities (57% mean percentage for faculty, 85%
for residents). Nine participants (13%; 7 faculty, 2 residents) had served on an IRB, and 13 (19%; 13 faculty, 0
residents) had served on a patient care ethics committee. A
smaller proportion of faculty (40%, n = 18) than residents
(64%, n = 18) had received formal ethics training during
their professional education (p = 0.03). Both faculty (60%,
n = 27) and residents (43%, n = 12) had encountered difficult ethical problems for which they felt unprepared in
clinical work. Overall, fewer faculty (13%, n = 6) and residents (4%, n = 1) had had this experience in the context of
research, although this may reflect relative inexperience in
research activities.
Importance of ethical issues (table 1). Item ratings
were initially subjected to a repeated measures multivariate analysis of variance (MANOVA) in which elements (12) was a within-subjects factor and respondent
(faculty vs. resident) and gender were between-subjects
factors. A main effect for elements (F( 11,57) = 34.0, p <
0.0001) showed that seven items were seen as very important (mean > 4.40), three items as important (3.00 > mean
> 4.10), and two items as less important (mean < 3.00). A
Discussion
This study characterizes the perceptions of 73 psychiatric
faculty and residents regarding the relative importance of
12 key elements in the ethical conduct of clinical research.
Three factors emerged in the responses: Safeguards and
Scientific Merit, Investigator Integrity and Relationship,
and Patient Benefit. The first factor, which was very
strongly endorsed, included regulated research requirements: informed consent, confidentiality protections, scientific considerations, and IRB processes. The second factor, also highly endorsed, captured elements related to
investigator professionalism: integrity, absence of conflicts of interest, trust, recruitment setting, and project
608
4.80
4.56
4.64
0.62
0.28
0.91
0.92
4.17
4.91
4.43
4.35
Factor 2: Investigator Integrity and
Relationship {a = 0.65)
A.
J.
0.46
4.37
1.17
2.46
1.09
0.59
0.66
Note.—LSD = least significant difference; SD = standard deviation.
1
Scale: 1 = not at all important, 5 = very important. A multivariate analysis of variance (MANOVA) on the 12 elements revealed a main effect for elements (p < 0.0001) and a main
effect for respondent—that is, faculty versus resident responses (p < 0.01). Similarly, a MANOVA on the three factors revealed a main effect for factors (p < 0.0001), a main effect for
respondent (p < 0.01), and an interaction of factors by respondent (p < 0.05). The above table was based on a sample size of 69 because of missing values for some items.
2
Cohen's d (the mean standardized difference between faculty and resident ratings). Differences between means are reliable by Fisher's LSD at a = 0.05 if:
1. Element overall means differ by at least 0.32.
• Within faculty, element means differ by at least 0.40.
• Within residents, element means differ by at least 0.53.
• Within each element, faculty versus resident means differ by at least 0.77.
2. Factor overall means differ by at least 0.20.
• Within each factor, faculty versus resident means differ by at least 0.40.
H.
G.
0.73
o
s
o
£
as
5'
Co
I
nt in Clinica
2.88
0.99
2.24
1.14
2.71
1.15
3.20
1.05
2.45
1.03
4.07
0.65
4.56
1.11
3.81
Likelihood of future benefit to patients
not involved in the project (X = 0.45)
Likelihood of immediate benefit to patients
involved in the project (X = 0.87)
Likelihood of immediate benefit to patients
not involved in the project (X = 0.88)
I.
0.80
0.90
3.08
0.80
3.55
0.86
2.83
Factor 3: Patient Benefit (a = 0.77)
0.32
1.18
3.44
1.11
3.68
1.21
3.30
K.
0.32
0.96
3.96
1.01
0.85
0.34
0.86
4.16
-0.34
0.15
0.35
0.34
0.83
0.57
0.02
0.82
4.45
4.87
4.48
4.24
4.45
0.70
0.41
0.65
0.91
4.46
3.85
F.
L.
Integrity of the investigator (X = 0.57)
Absence of conflicts of interest of
investigators (X = 0.61)
Trusting relationship between research
staff and patient participants (X = 0.38)
Recruitment setting of human subjects
for the project (X = 0.35)
Sources of funding for the project (X = 0.65)
0.78
4.44
C.
0.25
0.79
4.51
0.76
4.64
0.80
4.45
B.
0.26
0.41
0.26
d2
0.47
4.80
0.44
4.88
0.39
4.82
0.28
4.92
0.48
0.46
4.65
0.48
4.72
SD
Mean
Overall
(n = 71)
SD
Mean
Resident
(n = 25)
4.76
D.
4.76
E.
0.43
0.45
4.60
Factor 1: Safeguards and Scientific Merit (a = 0.68)
Adequacy of informed consent procedures with
human subjects (X = 0.69)
Efforts to protect the confidentiality of
participants in the project (X = 0.51)
Importance and scientific merit of the
project (x = 0.74)
Approval of the project by the institutional
review board (X = 0.51)
SD
Mean
Elements of ethical conduct
Faculty
(n = 46)
Table 1. Mean (SD) rated importance of 12 elements in the ethical conduct of clinical research1
IsEthi cally Im
Schizophrenia Bulletin, Vol. 29, No. 3, 2003
L.W. Roberts et al.
Figure 1. Perceived importance of safeguards and scientific merit
5 -,
Very
Important
2
— • — F a c u l t y (N • 46)
- -O- - Resident (N = 25)
d = 0.26
Mot at all
Important
1
Adequacy of informed
consent procedures with
human subjects
Efforts to protect the
confidentiality of
participants in the project
Importance & scientific
merit of the project
Approval of the project by
the Institutional Review
Board
Figure 2. Perceived importance of investigator integrity and relationship
Very
Important
S ,
*Q.
-•
2 J
— • — Faculty (N « 46)
- - O - Resident (N» 25)
d - 0.45
Not at all
Important
1
Integrity of the
investigator
Absence of conflicts of Trusting relationship Recruitment setting of Sources of funding for
interest of
between research staff human subjects for (he
the project
investigators
and patient
project
participants
Figure 3. Perceived importance of patient benefit
5 i
Very
Important
a.
D
— • — F a c u l t y (N • 46)
• o - Resident (N = 25)
d*0.80
Not at at
Ukelihood of FUTURE bened to
patients NOT knotted in the project
UketUwod of IMMEDIATE benefit to Likelihood of IMMEDIATE benefit to
patients involved in the project
patients NOT involved in the project
610
What Is Ethically Important in Clinical Research?
Schizophrenia Bulletin, Vol. 29, No. 3, 2003
funding sources. The third factor, which was viewed more
neutrally, comprised elements reflecting the value of current and future benefits to participants and to the broader
patient population.
Interestingly, residents rated the importance of each
factor, and nearly every element, more highly than did the
faculty members. On the other hand, a greater proportion
of faculty (57%) than residents (46%) reported encountering significant ethical dilemmas in their clinical work. The
main differences in perspective emerged with respect to
the role and weight of benefit. These two groups agreed on
relative ordering of the factors, and the single most important element for residents and faculty alike was investigator integrity.
perspectives, finding common ground as well as key differences between them, may assist academic departments
to approach their research activities with greater understanding and coherence.
Third, other key stakeholders in human research—
such as study participants, policy makers, institutional
leaders, and regulatory overseers—may value the elements differently. These attitudes, in turn, may give rise to
divergent attitudes and priorities. Further empirical
inquiry to discern the perspectives of diverse stakeholders
may be helpful in determining areas of overlap as well as
potential conflict. Because psychiatric research has met
with such extraordinary controversy over the past several
years, this evidence-based approach may help to accurately identify concerns, to gain a sense of proportion
regarding the topics of contention, and to minimize bias in
this highly sensitive area.
Fourth, these data underscore the need to better understand the ethical role of benefit in psychiatric studies. The
possibility of personal, familial, and/or societal benefit—
immediate and future—is known to be very important to
individuals who make the choice to enroll in human
research protocols (Sugarman et al. 1999). In addition, institutional review of psychiatric protocols often hinges upon
the potential risks and benefits for individual participants,
especially those from special populations (Dresser 1996;
Roberts 1999) Yet appropriate definitions of benefit have
not been developed to encompass the concepts of immediate, future, and indirect benefits for patient participants. Little attention has been given to psychosocial as well as biological dimensions of benefit and risk in human research.
For these reasons, our preliminary findings demonstrate the
need to study the relative contribution of participant benefit
in evaluating the ethics of human research protocols.
Implications. These pilot data have several implications
related to psychiatric research ethics. First, the data suggest
that psychiatrists place value on the duties of safeguarding
human subjects, performing high-quality scientific work,
and demonstrating respect for participants. Fostering trusting relationships and diminishing potential conflicts of interest in research were also affirmed in this study. The implication of this pattern is that psychiatrists accept responsibility
for these aspects of their role as professionals.
A related observation is that in light of the significance given to these ethical elements, it is critical for the
field of psychiatry to develop well-defined and evidenceanchored approaches to ethically important action in
human research. These approaches include decisions and
behaviors in implementing safeguards, evaluating scientific merit, building appropriate investigator-participant
relationships, and defining and managing conflicts of
interest, for which criteria are often implicit, intuitive,
and inconsistent within and across institutions (Levine
1986; Dresser 1996; Hoagwood et al. 1996; Spece et al.
1996).
Second, there was an interesting pattern of similarities
and differences in the responses by faculty and residents.
There was congruence in the perceptions of the relative
importance of the three factors and of key elements—for
example, researcher integrity. The residents in our sample
generally assigned greater importance to the elements than
did the faculty, and the reasons for this pattern were not
determined in this survey study. It may be that the intensive ethics focus throughout the professional training years
of these postgraduate physicians has had this impact. Or
the result may be attributable to other factors. It does not,
however, appear to be directly linked to the perceived frequency of complex ethical dilemmas, as these were more
commonly reported by faculty. Clearly, academically oriented training programs to undertake intensive dialogue
on human research ethics are needed. It may be discovered
that different points of view exist. Exploring divergent
Strengths and Limitations. This study is an initial effort
to assess the attitudes of psychiatrists regarding ethically
important elements in human research. It is only a small
contribution to the underdeveloped field of psychiatric
research ethics. Nevertheless, this project is a pilot selfreport study involving a small group of psychiatrists who
are primarily involved in clinical work and are located at
a single institution. While the items we assessed all may
be readily recognized as "socially desirable," we did find
significant differences in responses to the various items,
and the three distinct factors did emerge in our analysis.
Future empirical work will help to discern how psychiatrists would choose to "trade off' or "balance" different
positive values in making hypothetical decisions about the
ethical standing of specific human research activities.
Moreover, the extent to which these self-reported beliefs
reflect actual ethical behaviors of psychiatrists is a question for further study.
611
Schizophrenia Bulletin, Vol. 29, No. 3, 2003
L.W. Roberts et al.
Hoagwood, K.; Jensen, P.S.; and Fisher, C.B. Ethical Issues
in Mental Health Research With Children and Adolescents.
Mahwah, NJ: Lawrence Erlbaum and Associates, 1996.
Conclusion
This preliminary study addressed two questions: what
importance do psychiatrists assign to 12 elements contributing to the ethical caliber of clinical research, and do
psychiatry faculty view these issues differently than do
residents? We found that Safeguard and Scientific Merit
issues, as a set, emerged as the most salient considerations
for psychiatric faculty and residents alike. Residents overall assigned greater importance to the items than did the
faculty, and investigator integrity, adequacy of informed
consent procedures, and protection of confidentiality were
perceived by both groups as the most critical elements in
the ethical conduct of human research. Differences
emerged primarily in perceptions of the ethical relevance
of benefit.
It appears that psychiatrists in this sample perceived the
processes and professionals involved in clinical research—
that is, the "means" of human investigation—as central to its
ethical caliber. The "end" of patient benefit may be seen as
less important, from an ethical perspective, than the manner
in which this goal is obtained. These results are consistent
with the national regulatory movement toward a "culture of
accountability"—an approach that places tremendous ethical
responsibility upon academic institutions and individual
investigators who have the privilege of conducting research
with human participants.
In sum, this project is a reminder of the ethical
responsibilities accepted by investigators engaged in the
study of mental illness. Whether the patterns of response
we found are representative of psychiatrists in other academic and community settings remains uncertain. We
hope that future empirical studies will further clarify the
perspectives and priorities of different stakeholders,
enriching the understanding of the values and concerns of
those most centrally involved with psychiatric research.
The Human Radiation Experiments: Final Report of the
President's Advisory Committee. New York, NY: Oxford
University Press, 1996.
Hyman, S.E. Protecting patients, preserving progress:
Ethics in mental health illness research. Academic Medicine, 74(3):258-259, 1999.
Levine, R.J. Ethics and Regulation of Clinical Research.
Baltimore, MD: Urban and Schwartzenberg, 1986.
Levine, R.J. Ethics and Regulation of Clinical Research.
2nd ed. New Haven, CT: Yale University Press, 1988.
National Bioethics Advisory Commission. Research
Involving Subjects With Mental Disorders That May Affect
Decisionmaking Capacity. Rockville, MD: National
Bioethics Advisory Commission, 1998.
National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report: Ethical Principles and Guidelines for the
Protection of Human Subjects of Research. Washington,
DC: Government Printing Office, 1979.
Nunnally, J., and Bernstein, I. Psychometric Theory. New
York, NY: McGraw Hill, 1994.
Rappaport, M., and Carolla, B. NAMI publishes first-ever
consumer survey on scientific research. Individuals with
severe mental illnesses show strong support for research, but
want information and protection. Decade of the Brain, 1999.
Available at http://web.nami.org/pressroom/990430.html.
Roberts, L.W. Ethical dimensions of psychiatric research: A
constructive, criterion-based approach to protocol preparation. The Research Protocol Ethics Assessment Tool
(RePEAT). Biological Psychiatry, 46(8): 1106-1119,1999.
Roberts, L.W. Ethics and mental illness research. Psychiatric Clinics of North America, 25(3): 525-545, 2002.
Roberts, L.W.; Warner, T.D.; and Brody, J.L. Perspectives
of patients with schizophrenia and psychiatrists regarding
ethically important aspects of research participation.
American Journal of Psychiatry, 157(l):67-74, 2000.
Roberts, L.W.; Warner, T.D.; Brody, J.L.; Roberts, B.B.;
Lauriello, J.; and Lyketsos, C. Wash-outs and placebos:
Schizophrenia patients' and psychiatrists' assessments of
research protocols and participation decisions. American
Journal of Psychiatry, in press a.
Roberts, L.W.; Warner, T.D.; Nguyen, K.; Geppert, C ;
Rogers, M.; and Roberts, B.B. Schizophrenia patients' and
psychiatrists' perspectives on ethical aspects of symptom
reemergence during psychopharmacological research participation. Journal of Clinical Psychopharmacology, in
press b.
References
Brody, B. The Ethics of Biomedical Research: An International Perspective. New York, NY: Oxford University
Press, 1998.
Department of Health and Human Services, Office of Protection from Research Risks. Code of Federal Regulations,
Title 45: Public Welfare. Part 46: Protection of Human
Subjects, Regulation Governing Protections Afforded Children in Research (Subpart D: Additional Protections for
Children Involved as Subjects in Research). Washington,
DC: National Institutes of Health, 1991.
Dresser, R. Mentally disabled research subjects. The
enduring policy issues. JAMA, 276(l):67-72, 1996.
612
What Is Ethically Important in Clinical Research?
Schizophrenia Bulletin, Vol. 29, No. 3, 2003
Spece, R.; Shimm, D.; and Buchanan, A. Conflicts of
Interest in Clinical Practice and Research. New York, NY:
Oxford University Press, 1996.
ment Award (K02 MHO 1918-03) from the National Institute of Mental Health.
The contribution to this work made by Dr. Brian
Roberts was supported in part by the New Mexico VA
Health Care System, Albuquerque, NM.
Sugarman, J.; McCrory, D.C.; Powell, D.; Krasny, A.;
Adams, B.; Ball, E.; and Cassell, C. Empirical research on
informed consent. An annotated bibliography. Hastings
Center Report, 29(1):S1^2, 1999.
The Authors
Warner, T.W.; Roberts, L.W.; and Nguyen, K. How well
do psychiatrists understand the attitudes, motivations,
and experiences of schizophrenia protocol participants?
A pilot study comparing the predictions of 14 psychiatrists with the perspectives of 14 schizophrenia study
volunteers. Comprehensive Psychiatry, 44(3):227-233,
2003.
Laura Weiss Roberts, M.D., is Professor and Chair,
Department of Psychiatry and Behavioral Medicine,
Medical College of Wisconsin, Milwaukee, WI. Teddy
D. Warner, Ph.D., is Research Assistant Professor,
Department of Psychiatry, University of New Mexico
School of Medicine; and Interim Co-Director, University
of New Mexico Health Sciences Center Institute for
Ethics. Janet L. Brody, Ph.D., is Associate Scientist and
Clinical Director, Center for Family and Adolescent
Research, Oregon Research Institute. Khanh Nguyen,
M.S., is Credit Risk Analyst, Household Credit Services,
Las Vegas, NV. Brian B. Roberts, M.D., is President,
EthicsRx, LLC, Wauwatosa, WI.
Acknowledgments
This research was supported by a Young Investigator
Award to Dr. Laura Roberts from the National Alliance for
Research on Schizophrenia and Depression. Dr. Roberts is
also supported in part by a Research Scientist Develop-
613