Research on behavioral emergencies and crises
often involves participants who have suicidal tendencies, the potential
for violence, or a history of victimization or exposure to interpersonal
violence. Researchers who intentionally study participants with
issues such as these frequently confront ethical questions such
as: (1) will the research procedures or assessments (related to
suicidality, violence, or trauma) worsen the individual’s
condition and increase risk to self or others; (2) if a participant
becomes suicidal or potentially violent during his or her involvement
in the research protocol, will there be a need to break confidentiality
to protect the participant or others; and (3) if there are such
risks, how can the ethical obligations to the participant, others,
and the institution supporting the research can be met.
As Monahan, Appelbaum, Mulvey, Robbins, and Lidz (1993) have pointed
out, the APA Ethical Principles of Psychologists and Code of Conduct
(American Psychological Association, 2002), as well as other professional
codes of ethics, offer general guidance on ethical research practice,
but few practical guidelines for dealing with such specific research
issues. There has also been surprisingly little empirical study of
the questions involved.
Several years ago, for example, two of the authors
(PK & JK)
were required to address an Institutional Review Board (IRB) question
as to whether inquiring about the participant’s suicidal ideation,
and other variables associated with suicide risk, with a self-report
assessment instrument might stimulate or exacerbate self-destructive
impulses. The IRB’s concern about the possibility of negative
outcomes from participation was not assuaged by the reassurance that
clinical wisdom has long regarded as myth the thinking that asking
about suicidal ideation (as is routinely done in mental status exams)
will cause a patient to become more suicidal. As a result,
an extensive literature search was conducted with a focus on the
impact of participation in suicide–related research on suicidality,
and only one relevant study was found. Clum and Curtin (1993) had
investigated the reactivity of participants with chronic suicidal
ideation to self-monitoring of suicidal ideation. Participants were
required to rate the presence of suicidal ideation, its duration,
and the level of self-control they felt on a daily basis for 7 weeks.
During the course of this study, no subjects were lost to suicide
attempts or completions. Moreover, there was no (Contevidence that
increased attention to suicidal ideation resulted in an increase
in suicidality. To the contrary, there was some evidence of decreases
on measures of suicidal ideation following the period of self-monitoring,
something that could be interpreted as a positive effect of disclosure.
More recently, Oquendo, Stanley, Ellis, and Mann
(2004) described a randomized, controlled trial of pharmacotherapy
for suicidal behavior that attempted to minimize the risk of morbidity
and mortality. Lithium was compared to valproate for potential
antisuicidal effects in bipolar disorder patients. Several types of protective interventions
were included because the study participants had made past suicide
attempts and were considered at high risk of further suicidal behavior.
Thus, for example, so-called rescue medications could be added to
the treatment regimen if a participant became depressed or manic.
In addition, if a participant reported plans for a suicide attempt,
hospitalization, a change in medication, or increased monitoring
could be introduced. Moreover, the family of the participant was
involved in monitoring behaviors related to suicide risk and family
members were trained in interventions in the event of a relapse.
While these procedures should reduce the risk of harm to the participants,
the question is (as the authors have noted) whether they will also
decrease the incidence of suicidal behavior “in a way that
is inconsistent with ‘real world’ conditions” (p
1559) and, therefore, reduce the ecological validity of the study.
The outcome of this study will be closely watched.
Trauma researchers have given somewhat more attention
to the question of the impact of trauma research on participants.
Newman, Kaloupek, Keane, and Folstein (1997) expressed the opinion
that the trauma research community has failed to be self-reflective
and to examine the impact of trauma research on participants. Clearly,
inquiries about past traumatic events have been known to trigger
intense recollections, flashbacks, physiological reactivity, and
psychological distress, sometimes with concomitant suicidal or
homicidal impulses. Newman and her colleagues (Newman, Willard,
Sinclair, and Kaloupek, 2001) have therefore developed a questionnaire,
the Reactions to Research Participation Questionnaire (RRPQ), to
address this apparent shortcoming and to help promote ethical decision
making about research practice. In the initial testing of this
instrument, they found that participants’ appraisals
appeared to fall into five categories on factor analysis: personal
satisfaction, personal benefits, emotional reactions, perceived drawbacks,
and a global evaluation. Their plan is to continue to validate the
RRPQ as a means of empirically examining research participants’ distress
(as well as their perceived benefits and satisfaction) and as a way
of promoting the ethical treatment of those who participate in research.
Investigators who have studied the risk of violence
have also struggled with ethical questions about risk for participants
and others. In a large, prospective, multisite study (the MacArthur
Study of Mental Disorder and Violence), the research team worked
with a population of patients who had a relatively high potential
for becoming violent (Monahan, et al, 1993). They had ethical concerns
about the risk of violence for participants, the risk that others
in the community might be victimized by research participants,
and the risk that research staff themselves could be injured. Given
what they saw as a “paucity
of knowledge about the researcher’s duty to protect others
from harm” (p. 392), they allowed their different research
sites to develop their own methods for addressing these issues. Two
of the three sites interpreted their ethical obligations broadly.
Whenever routine data collection indicated the presence of suicidal
or homicidal ideation, they required the interviewers to explore
the degree of risk. Interviewers were provided with an outline of
areas into which they were to inquire, and they were instructed to
depart from the research protocol to do so. This information was
then brought to the senior staff and was reviewed by the principal
investigator and a clinical consultant. If the situation suggested
an imminent threat of serious harm to self or others, the senior
staff followed a set of guidelines and considered a range of preventive
The third site took a narrower view of their ethical obligations.
Interviewers were instructed to notify their supervisors if participants
behaved or were likely to behave in a manner that endangered their
own lives, or if they showed indications that they posed a substantial
threat to others. They were not, however, encouraged to deviate from
the research protocol to explore the potential for suicide or violence.
This group believed that it was beyond the abilities of non-clinically
trained research personnel to carry out such difficult and complicated
evaluations. This site saw their responsibility as reacting to manifest
threats of imminent harm, but not as obligating them to evaluate
more subtle risks to participants or others. If there appeared to
be imminent risk, the supervisory staff followed the same guidelines
as the other sites in considering preventive intervention.
Reports such as these, from very responsible investigators, indicate
that there is concern about the ethical obligations of the researcher
to protect high risk participants and third parties from harm. There
is little empirical data, however, on reactions to research participation.
Moreover, there is no consensus about the extent of the obligation
to detect potential for harm or exactly how the duty to protect can
best be discharged with research protocols that vary widely in form
and across contexts. Our hope in reviewing the efforts of those noted
above is to stimulate dialogue and further the investigation of these
issues, the resolution of which we believe to be of great importance
to the conduct of research on behavioral emergencies and crises.
If others have thoughts, comments, or information relevant to this
topic, we invite them to share it with us by contacting the first
author at: Phillip.Kleespies@med.va.gov
Psychological Association. (2002). Ethical principles of psychologists
and code of conduct. American Psychologist, 57, 1060-1073.
G., and Curtin, L. (1993). Validity and reactivity of a system of
self-monitoring suicide ideation. Journal of
Psychopathology and Behavioral Assessment, 15, 375-385.
J., Appelbaum, P., Mulvey, E., Robbins, P., and Lidz, C. (1993).
Ethical and legal duties in conducting research on violence: Lessons
from the MacArthur Risk Assessment Study. Violence
and Victims, 8,
E., Kaloupek, D., Keane, T., and Folstein, S. (1997). Ethical
issues in trauma research: The evolution of an empirical model for
decision making. In G. Kantor and J. Jasinski (Eds.): Out of the darkness:
Contemporary perspectives on family violence (pp. 271-281). Newbury
Park, CA: Sage.
E., Willard, T., Sinclair, R., and Kaloupek, D. (2001). Empirically
supported ethical research practice: The costs and benefits of research
from the participants’ view. Accountability
in Research, 8,
M., Stanley, B., Ellis, S., and Mann, J. J. (2004). Protection of
human subjects in intervention research for suicidal behavior. American
Journal of Psychiatry, 161, 1558-1563.