|
Prev Page 1 Next
VA COOPERATIVE STUDY #334:
I. SUMMARY OF FINDINGS
ON THE PSYCHOPHYSIOLOGICAL
ASSESSMENT OF PTSD
Terence M. Keane, PhD1
Danny G. Kaloupek, PhD1
Boston VA Medical Center and
Boston University School of Medicine
Lawrence C. Kolb, MD2
Sea Island, Georgia
Studies of the psychophysiology of PTSD can be traced to the observations of
Kardiner (1941), who viewed the psychological sequelae of combat as a ³physioneurosis.²
He characterized patients with this condition as having obvious elevations in
muscle tension, tachycardia, startle response, and a hyperresponsivity to external
stimulation. These observations led to empirical studies by Wenger (1948) and
then Dobbs and Wilson (1960), which found that physiological and psychophysiological
measures distinguished veterans with combat stress disorders from a variety
of comparison groups of veterans. The clinical observations of Kardiner and
the empirical studies of a variety of scientists, in-cluding Grinker and Spiegel
(1945), established the foundation for contemporary studies of the psycho-physiology
of PTSD among Vietnam veterans.
The purpose of this report is to update the field on the status of a multi-site
clinical trial that examined the ability of psychophysiological responses to
pre-dict PTSD diagnosis in a large sample of male Vietnam veterans. Data were
collected from 1990 to 1992 at 15 clinical research laboratories in VA Medi-cal
Centers in the United States. The scientific premise of the study came from
a series of publica-tions, reflecting work of three independent research groups,
that demonstrate how combat veterans with PTSD can be distinguished from combat
veterans without PTSD on the basis of psychophysiological information. A large-scale
study was needed to determine the generalizability of the accumulated findings
to a broader sample of help-seeking veter-ans. Such an effort would strengthen
the scientific foundation for this type of assessment and help clarify its diagnostic
implications for a larger seg-ment of the VA population.
Three studies conducted in the 1980s served as the contemporary foundation
for the multi-site project. The first of these, by Blanchard et al. (1982),
found that psychophysiological responses to stan-dardized sounds of combat discriminated
Vietnam veterans with PTSD from an age and gender matched comparison group.
At about the same time, Malloy et al. (1983) employed standardized audiovisual
cues of neutral and combat situations while measur-ing heart rate, skin resistance,
and subjective dis-tress. Male Vietnam veterans with PTSD were found to be more
physiologically responsive to the combat cues compared to veterans with psychiatric
impair-ment (but without PTSD) and veterans with no psychiatric impairment.
The third influential study was conducted by Pitman et al. (1987). These inves-tigators
adapted Peter Langıs (1985, 1995) imagery-based methods for studying the psychophysiology
of emotion and set about comparing Vietnam veter-ans with and without PTSD on
heart rate, skin conductance, and facial electromyographic re-sponses. The study
used a set of 30-second imagery scripts that were constructed in a systematic
fashion so that each contained stimulus, response, and mean-ing elements. Scripts
depicted individually-tailored traumatic and non-traumatic (e.g., neutral) experi-ences
that were compared to one another in terms of the physiological and subjective
reactions they pro-duced. Results based on this procedure demon-strated again
that veterans with PTSD react to indi-vidually- relevant combat cues with greater
arousal and distress than do veterans without PTSD.
These and other studies from a number of labora-tories provided impressive
preliminary evidence that individuals who qualify for a PTSD diagnosis can be
discriminated from those who do not meet PTSD criteria when the groups are compared
on their psychophysiological response to challenge tasks that involve presentation
of trauma-relevant cues (see Orr & Kaloupek, 1997; Prins et al., 1995). Despite
the impressive consistency of findings, there are methodological limitations
that raise questions about the individual estimates of diagnostic accu-racy
obtained by any particular study in the set. These limitations include: (a)
problems of inflated base rates due to the fact that PTSD patients typi-cally
comprised 1/3 to 1/2 of the study sample, a higher rate than is likely to occur
in the population at large; (b) several of the comparison groups con-sisted
of veterans who were not seeking help for their problems (e.g., VA health services)
while par-ticipants in the PTSD group were; (c) only two studies (Blanchard
et al., 1991; Orr et. al., 1993) attempted to cross-validate findings within
the con-text of a single study; and (d) most of the studies did not employ contemporary
multivariate data ana-lytic strategies to investigate classification accuracy.
Many of the studies conducted during the 1980s
Prev Page 2 Next
found classification rates of 75-95% when psychophysi-
ological variables were used for post hoc classification
relative to interview-based PTSD diagnosis. This places
psychophysiological responses to a trauma-relevant chal-
lenge task among the most reliable and robust physiologi-
cal indicators of any psychological disorder or condition.
This high level of promise, coupled with the previously
noted methodological limitations, provided justification
for a more definitive study that would collect data from a
large sample of subjects in order to test the utility of
psychophysiological reactivity as an index of PTSD. A
proposal was submitted to the VA Cooperative Studies
Program by principal investigators Terence Keane and
Lawrence Kolb because it was recognized that a suitably
large sample of individuals with PTSD would be difficult
to accumulate at one site within a reasonable time period.
The Cooperative Studies research operation was well suited
to the task because it specializes in multi-site investigations
that capitalize on the exceptional research talent and large
patient populations available across the VA system.
In time, the proposal successfully negotiated the rigors
of the Cooperative Studies Program, and Cooperative
Study #334 was initiated. The goals of the project included
conducting a comprehensive utility analysis of psycho-
physiological variables in order to generate indices of
sensitivity, specificity, predictive power of a positive test,
and predictive power of a negative test. These components
of a psychometric analysis are typically used in the devel-
opment of any new psychological or medical test. As a set,
they provide a comprehensive understanding of the accu-
racy associated with a particular measurement tool.
Method. CS #334 was directed from the Chair's office at
the Boston VAMC, with operational support from the
Cooperative Studies Program Coordinating Center in
Menlo Park, CA. Fifteen sites were selected in a competi-
tion involving forty-three medical centers that submitted
full applications expressing interest in serving as a partici-
pating site. Training for the study included a centralized
workshop for study clinicians devoted to administration
of diagnostic interviews, and a similar workshop for study
technicians concerning the laboratory procedures and data
management. The workshops were supplemented by on-
site consultation with an expert on psychophysiological
measurement, as well as regularly scheduled telephone
conference calls involving all site investigators and project
directors. Ongoing telephone supervision and consulta-
tion with experts was routinely available for any problems
that arose after the study was initiated. In addition, an
annual meeting of project staff from all fifteen sites was
organized to present updated information and to resolve
collective difficulties that sites identified.
Participants. Primary recruitment sources for the project
included VA Medical Center inpatient and outpatient pro-
grams in psychiatry, substance abuse, and PTSD. All par-
ticipants were male, had served in the Vietnam theater of
operations between August 1964 and May 1975, and were
receiving some form of service from DVA. Individuals
were excluded from participation if they experienced car-
diovascular disease including stroke, myocardial infarc-
tion, angina, or uncontrolled hypertension. Individuals
also were excluded if they were diagnosed with an endo-
crine disorder, seizure disorder, or organic brain syn-
drome. Participants were required to be free of psychotro-
pic medications or beta-andrenergic blocking medication,
though individuals who were taking such medications at
the time of screening remained eligible if they and their
physician agreed to medication withdrawal through the
period of the study. Acute alcohol use and illicit drug use
during the course of the study were restricted, and urine
samples were obtained to ensure compliance with this
regulation. A total of 2,115 individuals were screened as
potential participants, and 654 were excluded on the basis
of factors listed above. Of the 1,461 veterans eligible for
participation, 220 terminated prior to the psychophysi-
ological testing, including 133 who did so before complet-
ing the diagnostic interview for PTSD.
Psychological Assessment. Because there is no absolute
criterion, no "gold" standard, for the measurement and
diagnosis of PTSD, a decision had to be made about how
best to establish the diagnosis for the purposes of the study.
The instrument selected for this purpose was the PTSD
module of the Structured Clinical Interview for the DSM-
III-R (SCID; Spitzer et al., 1989), arguably the most widely
used PTSD interview format available at the time. Positive
features of the module included its administration by
clinically trained individuals, the available pool of VA
clinicians who had experience using the interview with
PTSD patients, and the fact that the module was part of a
comprehensive instrument that could be used for other
diagnostic determinations in the study.
During actual data collection, the War Stress Interview
(Rosenheck & Fontana, 1989) was administered first to
obtain sociodemographic, psychosocial history, and men-
tal health information. Participants next completed the
SCID, including modules for major depression, bipolar
disorder, schizophrenia, alcohol abuse and drug abuse,
panic disorder, social phobia, obsessive-compulsive disor-
der, dissociative disorder, and PTSD (combat-related and
non-combat related). Antisocial and borderline personal-
ity disorders were assessed using modules of the SCID-II.
All interviews were audiotaped, and a selected subset was
employed in a study of interater reliability.
The SCID interview was completed by 1,328 participants
who were assigned to the following groups: Current PTSD
(n = 778), Lifetime PTSD (n = 181) and Never PTSD (n =
369). After the WSI and SCID, participants completed the
MMPI-2, which contains the Keane PTSD scale (Keane et
al., 1984), as well as the Combat Exposure Scale (Keane et
al., 1989), the Mississippi Scale for Combat Related PTSD
(Keane et al., 1988), and the Laufer Parsons Inventory
(Laufer et al., 1981) to assess combat-related guilt.
Psychophysiological Assessment. The study applied both
the standardized audiovisual format developed by Malloy
et al. (1983) and the idiographic imagery format used by
Pitman et al. (1987). Psychophysiologic measures included
heart rate, skin conductance, forehead electromyogram,
Prev Page 3 Next
and systolic and diastolic blood pressure.
The procedures for all sites were identical and standard-
ized from minute to minute throughout the course of the
laboratory tasks. All participants were introduced to the
psychophysiological laboratory on a separate day after
completing both the diagnostic interviewing and psycho-
logical assessment. They were first oriented to the labora-
tory before sensors were attached for physiological record-
ing. At the start of the procedure, each participant was
asked to relax quietly for a period of 10 minutes. Next, they
engaged in mental arithmetic, a generic stressor task, for 2
minutes before sitting quietly for a 5-minute rest period.
The next phase of the laboratory procedure included pre-
sentation of a set of 6 standardized neutral audiovisual
presentations that lasted 1 minute each. These consisted of
6 still pictures and an accompanying sound track pre-
sented for 1 minute each. Physiological measures were
recorded during each presentation and subjective ratings
of distress (SUDS) were obtained during a 30 second
period between presentations. The neutral presentations
were followed by a 5-minute rest period, which was fol-
lowed by presentation of 6 combat audiovisual scenes
according to the same procedure used for the neutral
scenes. The combat scenes were followed by a 5-minute
recovery period.
Following another 5-minute rest period, participants
were presented a set of 4 tape-recorded imagery scripts,
each of which consisted of 4 sequential 30-second periods:
the baseline, reading of the script, imagining the script, and
recovery. Two of the imagery scripts (the first and third in
the sequence) had neutral content that was standardized
across subjects. The other two scripts depicted the two
most upsetting or stressful combat experiences identified
by the participant during a special interview conducted by
the study clinician in conjunction with the diagnostic inter-
viewing. A 5-minute recovery period concluded the psy-
chophysiological assessment procedure.
Most participants were debriefed about the study-not
just the laboratory session-after the recording sensors
were removed. However, a subset of approximately 25% of
the participants was randomly selected for a second labo-
ratory testing that was scheduled on a separate day to
examine the stability or test-retest reliability of the physi-
ological assessment procedures. Debriefing for these par-
ticipants followed the second challenge test session.
Primary data analysis for the study used logistic regres-
sion to predict membership in Current PTSD or Never
PTSD groups. The total study group with completed psy-
chophysiological testing (n = 1,241) was divided randomly
into two samples so that the resulting prediction equation
could be cross-validated. The calibration sample that gen-
erated the equation was comprised of approximately two-
thirds (n = 740) of the participants. The validation sample
to which the equation was re-applied was comprised of the
remaining one-third (n = 371).
Findings. Results of this multi-site investigation provide
definitive support for relatively heightened physiological
responsivity to combat cues for male veterans with a PTSD
diagnosis. The response differences are, predictably, great-
est between the Current PTSD and Never PTSD groups.
The differences are found in conjunction with both the
standardized audiovisual format and the idiographic im-
agery format of challenge testing.
In addition, higher absolute levels of physiological mea-
sures were found at baseline for the Current PTSD group.
These differences specifically involve heart rate and skin
conductance measures and suggest that individuals with
current PTSD are more aroused at rest than those in the
other two groups. Similar differences have been reported
by several of the previous small-scale studies, and we
(Prins et al., 1995) have interpreted them as an indication of
anticipatory fear triggered in veterans with PTSD who find
novel environments such as a psychophysiology labora-
tory to be threatening. A recent study by Orr and his
colleagues (1998) adds to the evidence that veterans with
PTSD do not show differences in resting physiological
measures when they are at home, in a presumably non-
threatening environment. Such evidence argues against
the notion of fixed physiological elevations among indi-
viduals with PTSD and is consistent with situational influ-
ences on baseline arousal such as anticipatory anxiety.
It was evident that the absolute level of responding to
trauma-related cues was lower in this study than in many
of the prior studies. This comparative reduction may well
be attributable to the application of stringent inclusion and
exclusion criteria that limited participation to relatively
healthy subjects who were willing and able to come off
their psychotropic medications.
Some findings relevant to the magnitude of responding
by the study participants come from follow-up analysis of
responders and non-responders within the current PTSD
group. This analysis demonstrates that the most respon-
sive individuals show heart rate increases of 6-7 beats per
minute, consistent with values obtained in the earliest
studies. These Responders also displayed selectively greater
symptoms of PTSD, and more symptoms of war-related
guilt and depression on self-report scales. It is quite pos-
sible that application of the inclusion/exclusion criteria
caused the study to enroll a broader range of PTSD severity
than had the preliminary studies in the field, particularly
at the lower end of the severity spectrum.
Using physiological variables alone, the optimal per-
centage for correct classification was 69% in the calibration
sample and 64% in the validation sample, yielding a com-
bined rate of approximately 67%. These results are lower
than those previously reported, but some of this reduction
was expected because the sampling strategy differed from
past efforts by emphasizing the absence of medications or
drugs as potential influences on the physiological vari-
ables. Nonetheless, the results are quite respectable for a
biological test of a psychological condition, surpassing classi-
fication rates observed in other such arenas (e.g., the Dex-
amethasone Suppression Test for Depression; Insel &
Goodwin, 1983).
In terms of the participants who met criteria for a lifetime
PTSD diagnosis, we initially predicted that they would
Prev Page 4 Next
show physiological reactivity that ranged between that
shown by the Current PTSD and Never PTSD groups. This
was generally the case. For the test-retest reliability compo-
nent of the project, we learned that over a period of ap-
proximately 10 days there was a significant reduction in
responsiveness overall, but the level of consistency (e.g.,
relative rank for an individual) across the two assessments
was quite high.
In conclusion, CS #334 provides strong empirical sup-
port for the presence of objectively measured psycho-
physiological reactivity to trauma cues as a distinguishing
feature of PTSD. The findings suggest that individuals
with the strongest physiological responses are the most
impaired on clinical self-rating scales, and they endorsed
more symptoms of war-related guilt and depression. Fi-
nally, the findings from the array of psychological tests,
questionnaires, and interviews employed in the study
indicate that Vietnam veterans with current PTSD con-
tinue to suffer from a broad range of psychological symp-
toms, multiple comorbid conditions, marital and family
dysfunction, vocational impairment, financial instability,
and other psychosocial difficulties. The extent to which
these findings regarding psychophysiological and cardiac
reactivity can predict service utilization, the development
of additional health problems, and continued psychologi-
cal distress remains an important empirical question that
warrants additional study.
REFERENCES
BLANCHARD, E.B., KOLB, L.C., PALLMEYER, T.P., &
GERARDI, R.J. (1982). A psychophysiological study of
posttraumatic stress disorder in Vietnam veterans. Psychiatric
Quarterly, 54, 220-229.
BLANCHARD, E.B., KOLB, L.C., & PRINS, A. (1991). Psycho-
physiological responses in the diagnosis of posttraumatic stress
disorder in Vietnam veterans. Journal of Nervous and Mental
Disease, 179, 99-103.
DOBBS, D. & WILSON, W.P. (1960). Observations on persis-
tence of war neurosis. Diseases of the Nervous System, 21, 686-691.
GRINKER, R.R. & SPIEGEL, J.P. (1945). Men under stress. Phila-
delphia: Blakiston.
INSEL, T.R. & GOODWIN, F.K. (1983). The dexamethasone
suppression test: Promises and problems of diagnostic labora-
tory tests in psychiatry. Hospital and Community Psychiatry, 34,
1131-1138.
KARDINER, A. (1941). The traumatic neuroses of war. New York:
Hoeber.
KEANE, T.M., CADDELL, J.M., & TAYLOR, K.L. (1988). Mis-
sissippi Scale for Combat-Related Posttraumatic Stress
Disorder:Three studies in reliability and validity. Journal of
Consulting and Clinical Psychology, 56, 85-90.
KEANE, T.M., FAIRBANK, J.A., CADDELL, J. M., ZIMERING,
R. T., TAYLOR, K. L., & MORA, C. A. (1989). Clinical evaluation
of a measure to assess combat exposure. Psychological Assess-
ment, 1, 53-55.
KEANE, T.M., MALLOY, P.F., & FAIRBANK, J.A. (1984).
Empirical development of an MMPI subscale for the assess-
ment of combat-related posttraumatic stress disorder. Journal of
Consulting and Clinical Psychology, 52, 888-891.
LANG, P.J. (1985). The cognitive psychophysiology of emo-
tion: Fear and anxiety. In A. H. Tuma & J.D. Maser (Eds.), Anxiety
and the anxiety disorders (pp. 131-170). Hillsdale, NJ: Erlbaum.
LANG, P.J. (1995). The emotion probe: Studies of motivation
and attention. American Psychologist, 50, 372-385.
LAUFER, R.S., YAGER, T., FREY-WOUTERS, E., &
DONNELLAN, J. (1981). Legacies of Vietnam, vol. III: Post-war
trauma: Social and psychological problems of Vietnam veterans in the
aftermath of the Vietnam War. Washington, D.C.: U.S. Government
Printing Office.
MALLOY, P.F., FAIRBANK, J.A., & KEANE, T.M. (1983).
Validation of a multimethod assessment of posttraumatic stress
disorders in Vietnam veterans. Journal of Consulting and Clinical
Psychology, 51, 488-494.
ORR, S.P. & KALOUPEK, D.G. (1997). Psychophysiological
assessment of posttraumatic stress disorder. In J.P. WILSON, &
T.M. KEANE (Eds.), Assessing psychological trauma and PTSD: a
handbook for practitioners (pp. 69-97). New York: Guilford.
ORR, S.P., MEYERHOFF, J.L., EDWARDS, J.V., & PITMAN,
R.K. (1998). Heart rate and blood pressure resting levels and
responses to generic stressors in Vietnam veterans with
posttraumatic stress disorder. Journal of Traumatic Stress 11, 155-
164.
ORR, S.P., PITMAN, R.K., LASKO, N.B., & HERZ, L.R. (1993).
Psychophysiological assessment of posttraumatic stress disor-
der imagery in World War II and Korean combat veterans.
Journal of Abnormal Psychology, 102, 152-159.
PITMAN, R.K., ORR, S.P., FORGUE, D.F., DE JONG, J.B., &
CLAIBORN, J. M. (1987). Psychophysiologic assessment of post-
traumatic stress disorder imagery in Vietnam combat veterans.
Archives of General Psychiatry, 44, 970-975.
PRINS, A., KALOUPEK, D.G., & KEANE, T.M. (1995). Psycho-
physiological evidence for autonomic arousal and startle in
traumatized adult populations. In M. J. Friedman, D. S. Charney
& A. Y. Deutch (Eds.), Neurobiological and clinical consequences of
stress: from normal adaptation to post-traumatic stress disorder (pp.
291-314). Philadelphia: Lippincott-Raven.
ROSENHECK, R. & FONTANA, A. (1989). War Stress Interview
(WSI). West Haven, CT: Department of Veterans Affairs, North-
east Program Evaluation Center.
SPITZER, R.L., WILLIAMS, J. B.W., GIBBON, M., & FIRST, M.
D. (1989). Structured Clinical Interview for DSM-III-R-patient ver-
sion (SCID-P). New York: New York State Psychiatric Institute,
Biometrics Research Division.
WENGER, M.A. (1948). Studies of autonomic balance in
army air forces personnel. Comparative Psychology Monographs,
19.
VA COOPERATIVE STUDY #334:
II. OVERVIEW OF THE PLANNING
PROCESS
Many investigators feel that the Cooperative Studies Pro-
gram of the Office of Research and Development in the
Department of Veterans Affairs is the crown jewel in an
extraordinarily productive research program. The Coopera-
tive Studies Program capitalizes on the exceptional pool of
professional talent associated with VA research and clinical
programs, and the large number and wide geographic distri-
bution of VA facilities as a vehicle for conducting multi-site
clinical trials. The tremendous success and broad impact of
the program on health care is evident in a roster of studies
that includes monumental investigations such as those ex-
amining the effects of aspirin on heart disease and testing
treatments for hypertension.
Prev Page 5 Next
Not surprisingly, the first step in developing a Coopera-
tive Study proposal involves gaining familiarity with the
practices and procedures for accessing the program. While
this step can appear daunting for investigators accus-
tomed to individual VA Merit Review or NIH grant appli-
cations, the logic of the program requirements becomes
apparent when one considers the need to manage the
complexity of a large research enterprise and the cost of a
multi-site initiative. Mechanisms are in place to insure that
the proposal is feasible in terms of coordination and that it
is likely to produce findings that justify the costs involved.
With these considerations in mind, Cooperative Study
#334 was based on an idea supported by a substantial body
of evidence drawn from a number of projects conducted by
individual investigators over nearly a decade.
The first formal stage in a Cooperative Study proposal is
submission of a Planning Request. This document is simi-
lar to the brief Letter of Intent that often precedes proposal
submission for individual grant programs. The principal
proponent for the study submits a ten-page request that
contains the objectives of proposed research, the impor-
tance of the topic to the VA, justification and need for
multi-site study, a review of the literature, and a brief
description of the proposed study design. The description
of the study design need not contain all details of the
proposed study. Indeed, this study description aims to
demonstrate to reviewers the logical integration of the
questions posed, the design considered, the data to be
collected, and the outcomes to be measured, as well as the
importance of the questions in clinical and, especially,
scientific context.
The Planning Request is submitted to the Chief Research
and Development Officer through the local VA Research
and Development Office. Approval of the plan requires
favorable evaluation from the CSP review panel. If ap-
proved, the proposal is assigned to a specific coordinating
center, a study biostatistician is allocated to the project, and
detailed planning begins. CS #334 had the good fortune to
be assigned to the Palo Alto center, under the direction of
Kenneth James, Ph.D. Dr. James, in turn, was responsible
for the assignment of the project biostatistician, Ronald
Thomas, Ph.D.
Planning and developing a multi-site study involves
many different individuals in addition to the Principal
Proponent and study biostatistician. Planning committee
members are solicited by the Principal Proponent and
nominated with the approval of the CSP. Our planning
committee for CS #334 included Edward Blanchard, Ph.D.,
SUNY-Albany, Roger Pitman, M.D., and Scott Orr, Ph.D.,
of the Manchester VA and Harvard Medical School, and
Patrick Boudewyns, Ph.D., of the Augusta VA, in addition
to ourselves.
The Principal Proponent for the study assumes leader-
ship in the planning process and is actively involved in the
nomination and selection of members of the planning
committee. He or she serves as the chair for the planning
meetings and has primary responsibility for coordination
and writing of the full proposal for submission. Unlike
other grants that are submitted to funding agencies exclu-
sively as paper documents, the Cooperative Studies Pro-
gram offers the Principal Proponent and the study biostat-
istician the opportunity to defend their proposal orally
before the Cooperative Studies Evaluation Committee
(CSEC). This oral presentation gives the Principal Propo-
nent a forum in which to emphasize details that may be of
special importance and to answer questions that members
of the CSEC have regarding the application. The biostatis-
tician is likely to be called upon to defend the choice of
study design and data analytic strategies. Without ques-
tion this is one of the more challenging experiences of
heading a Cooperative Study.
The real excitement begins when a Cooperative Study
proposal finishes the review process with highly favorable
ratings and is designated to receive funding.
Typically, a study is funded to establish and maintain a
Study Chair's office, a local entity under the direction of the
Principal Proponent that serves to monitor and adminis-
trate the conduct of the study itself. Our Study Administra-
tor was Adele George, R.N., an experienced research nurse.
One of her initial tasks was to work closely with the Study
Chair and the biostatistician to solicit study-site investiga-
tors. For CS #334, forty-five applications were submitted
by investigators from across the country, with fifteen sites
ultimately selected to receive funding to participate. Im-
portant considerations in these decisions were the research
accomplishments of the potential site investigator, the
availability of suitable sources for subject recruitment at a
site, and contribution to the geographic mix of the study.
The next phase of the Cooperative Study is to provide
education about and training in the research protocol and
procedures so that staff at all participating sites are admin-
istering the protocol in as near an identical fashion as
possible. Our training program consisted of two phases.
First, we convened a three-day meeting in New Orleans
to train staff from all sites, particularly the study clinicians,
in the details of participant selection and screening. As part
of the training, research technicians were taught the use of
the psychophysiological equipment in the context of the
protocol we had designed. The second phase involved on-
site training at which Danny Kaloupek or Scott Orr set up
each laboratory and reviewed the study methods with the
local study staff. Throughout the study, Adele George
provided daily supervision of all 15 study sites, answering
questions and resolving problems in a manner that aimed
to maintain a high level of procedural consistency and
integrity.
The Executive Committee of a Cooperative Study al-
ways includes the Principal Proponent and the Study
Biostatistician as well as a small group of investigators who
are actively involved in the research study either as partici-
pating investigators or specialized consultants. The Execu-
tive Committee is the overall managing body that makes
decisions for operational components of the study. Any
changes in protocols are made by the Executive Commit-
tee, and this group typically has the final word on publica-
tions regarding study data. For CS #334, many of the
Prev Page 6 Next
members of the Planning Committee also agreed to serve
on the Executive Committee.
The Data Monitoring Board (DMB) consists of experts in
the subject matter of the project as well as external biostat-
isticians and other personnel with appropriate scientific
skills. A DMB is responsible for monitoring, evaluating,
and making recommendations regarding aspects of the
ongoing study. In contrast to the Executive Committee,
which is blinded with regard to the interim data analyses,
DMB members are informed about interim findings as the
study progresses. This oversight is particularly important
when clinical procedures with life and death implications
are involved. It evaluates at periodic intervals whether the
study should continue, assesses the performance of each
data collection site, makes recommendations regarding
their continued participation, and provides recommenda-
tions to the Executive Committee regarding changes in the
protocol. The Study Chair nominates candidates for the
Board, but members are selected by the CSP personnel.
Among the challenging aspects of conducting our Coop-
erative Study was gaining approval from the 15 Human
Studies Committees (HSM) at participating VA Medical
Centers, as well as the HSC for the Cooperative Studies
Program itself. This latter committee had overarching re-
sponsibility for determining that the protocol was safe and
consistent with the interests and welfare of participants.
However, it was also necessary to work with the local HSC
and to address their individual questions and concerns.
We were fortunate to have site investigators who already
had excellent relationships with their local HSC and were
able to resolve this complex set of dictates in a manner that
allowed a uniform protocol to be implemented.
The CSP Human Studies Committee met with the Data
Monitoring Board annually during the study to follow the
course of data collection. While CS #334 did not encounter
difficulties in this regard, it is possible for the HSC to
recommend changes in the protocol because of problems
that emerge during implementation, or to even terminate
a study if participants' well-being is compromised.
It should be apparent that completing a successful Coop-
erative Study is a complex process involving individuals at
the study sites, the Coordinating Center, and the Chair's
office. In addition, the DMB and the HSC work with the
Executive Committee to identify problems that might un-
dermine successful study completion. Difficulties at any
site or with the protocol as a whole are managed immedi-
ately by one or another of these Committees.
Most Cooperative Studies take several years to com-
plete. As a result, there are often personnel changes at
study sites. In the case of CS #334, these changes included
site investigators, study clinicians, and study technicians.
Our study also had the less typical experience of having
personnel changes in the CSP office itself. Along the way,
Philip Lavori, Ph.D., became the director of the Palo Alto
center and Frank Hsieh, Ph.D., became the study biostatis-
tician. It is a testament to the professionalism of both
original and new CSP staff that the study was able to
succeed despite these major transitions.
In summary, the VA Cooperative Studies Program of-
fers a unique opportunity to address large-scale questions
in a system that shares many common characteristics in
terms of patients, staffing, and service delivery. Partici-
pating in a Cooperative Study at any level provides the
investigator with a perspective on health care and on
science that is otherwise difficult to attain. Many of the
investigators who were actively involved in the Coopera-
tive Study on the psychophysiology of PTSD stated forth-
rightly that the experience was among the best scientific
efforts with which they had been involved over the course
of their careers. Thus, while the process is complex, the
rewards of research within the Cooperative Studies frame-
work are many, both for the individuals involved and for
the body of scientific knowledge to which they are contrib-
uting with their efforts.
Looking ahead, Cooperative Study #420 is a multi-site trial
investigating group therapy in PTSD. It is chaired by Mat-
thew Friedman, M.D., Ph.D., and Paula Schnurr, Ph.D., of the
National Center for PTSD in White River Junction, VT. These
investigators have successfully launched the largest treat-
ment study of PTSD proposed to date and perhaps one of the
most important examinations of group therapy in the treat-
ment of any psychological condition. We wish the propo-
nents and the study site investigators the best of luck in
completing their clinical trial.
ABOUT THE VA COOPERATIVE
STUDIES PROGRAM
Cooperative studies first began in the VA system in 1946,
with landmark research in the treatment of tuberculosis. This
research established a framework that evolved into the present-
day Cooperative Studies Program. The Cooperative Studies
Program (CSP) was established as a division of the Medical
Research Service in 1972. Traditionally, it has coordinated
multi-center clinical trials of new therapies or new uses for
standard treatments. In 1990, a program to facilitate multi-site
health services research, Cooperative Studies in Health Ser-
vices, was created within the Health Services Research and
Development Service. In 1996, these programs were merged,
and the Cooperative Studies Program became the fourth ser-
vice in the VA's Office of Research and Development. The
program's mission now encompasses all fields of research
important to veteran's health care: medical research, health
services research, and rehabilitation research.
CSP utilizes the power of multi-center studies to achieve
more definitive findings than might be available in single-site
studies. With its many hospitals and integrated networks, the
Veterans Health Administration is an ideal place to conduct
large-scale cooperative research. Such work has a direct im-
pact on veterans' clinical care, and provides a national re-
source to the health-care community within the VA and be-
yond.
For more information about the Cooperative Studies Pro-
gram, please contact: Cooperative Studies Program
Department of Veterans Affairs (121D)
810 Vermont Avenue, NW
Washington, DC 20420
(202) 273-8229
ping.huang@mail.va.gov
Prev Page 7 Next
RESEARCH AT THE NATIONAL CENTER FOR PTSD:
CLINICAL LABORATORY AND EDUCATION DIVISION
Kent D. Drescher, PhD
The National Center for PTSD Clinical Laboratory in
Menlo Park, California, provides and studies both inpa-
tient and outpatient clinical services. Currently there are
men's (48-bed) and women's (10-bed) inpatient programs.
These programs provide intensive 45-day multi-modal
inpatient care. Group psychotherapy focuses on develop-
ing more successful here-and-now coping skills and life
plans, and typically includes group trauma focus. In the
outpatient setting, the PTSD Clinical Team (PCT) staff
provide outpatient assessment and treatment at four loca-
tions: Menlo Park, San Jose, Santa Cruz, and Monterey.
Dr. Kent Drescher is currently collecting data for an
inpatient program evaluation protocol. Previous research
by Ford (1995) found clinician ratings of patient object-
relations to predict treatment outcome. The present study
is an attempt to replicate and extend this finding. PTSD and
associated symptom severity, social support, coping, and
quality of life are being monitored. All patients are as-
sessed at admission, discharge, and follow-up. Approxi-
mately 40 patients have completed the protocol. Dr. Ron
Murphy is collecting pilot data on the effectiveness of a
group treatment designed to increase awareness of and
motivation to change PTSD-related problem behaviors.
Early data indicates that many patients do not perceive a
need to change various PTSD symptoms, especially anger
and isolation. Non-confrontational motivational interview-
ing techniques appear useful in enhancing patient aware-
ness of the need to change various problem behaviors. It is
hoped that post-treatment outcome for PTSD patients may
improve with increased patient willingness to change prob-
lem behaviors. Dr. Murphy is also completing a manu-
script describing findings from an NIAAA-funded study
of the effects of combat exposure and childhood trauma on
alcohol use. Results suggest that a history of severe physi-
cal punishment as a child correlates with higher combat
exposure. Also, compared to childhood trauma, combat
exposure is a better predictor of past-year alcohol con-
sumption. Finally, IRB approval has been received for
analysis of findings related to clinical assessment data
collected on patients who received inpatient treatment
between 1990 and 1996. Dr. Drescher and other NCPTSD
staff plan a series of articles to describe these data. Data-
base and staffing infrastructure have been developed to
begin efficiently evaluating outcomes for outpatients treated
by our PCT program at all four sites. We hope to submit a
protocol shortly to begin monitoring outcomes in this
setting.
Dr. Steve Woodward directs the NCPTSD Sleep Labora-
tory. This lab is in year two of a three-year Merit Review
Grant designed to record PTSD nightmares and to assess
the physiologic background of waking intrusive phenom-
ena such as intrusive thoughts and flashbacks. In this
context he has initiated 24-hour ambulatory monitoring in
PTSD inpatients. This project will also provide precise
estimates of the contribution of physical activity to 24-hour
integrated urinary norepinephrine, epinephrine, and cor-
tisol. This laboratory is engaged in the development of two
advanced diagnostic methodologies for application to
PTSD. The first is an enhanced trauma cue reactivity pro-
tocol which assesses patients' arousal responses to visual
and auditory reminders of their traumatic experiences.
The second is an auditory version of the emotional Stroop
paradigm, which assesses the degree to which trauma-
related cues deflect a patient's attentional resources from a
primary task. Additionally, the lab is involved in a formal
collaboration with Richard J. Ross's NIH-funded examina-
tion of sleep in monozygotic twins discordant for PTSD.
Dr. Annabel Prins is involved in the development and
validation of a 4-item PTSD screening instrument for use in
primary-care settings. Currently data are being collected in
two primary-care clinics. The protocol involves adminis-
tration of the PTSD screen followed by a CAPS interview.
Preliminary data indicate that a high percentage of veter-
ans in primary-care clinics have trauma exposure. The
screen instrument appears to have good internal consis-
tency, adequate test-retest reliability, and adequate con-
current validity with the CAPS interview. To date, 29% of
veterans screened meet criteria for full PTSD, and 48%
have partial PTSD symptoms. In addition, Dr. Prins's other
research projects include survey research on the reasons
and consequences of trauma concealment, laboratory re-
search on possible moderators of psychophysiological re-
sponding in PTSD, and archival research on lipid levels in
women with and without PTSD.
Selected Bibliography
ABUEG, F.R., DRESCHER, K.D., & KUBANY, E.S. (1994). Natu-
ral disasters. In F.M. Dattilio & A. M. Freeman (Eds.), Cognitive-
behavioral strategies in crisis intervention (pp. 238-257). NY: Guilford
Press.
GUSMAN, F.D., STEWART, J., YOUNG, B.H., RINEY, S.J.,
ABUEG, F.R., & BLAKE, D.D. (1996). A multicultural develop-
mental approach for treating trauma. In A.J. Marsella, M.J. Fried-
man, E.T. Gerrity, & R.M. Scurfield (Eds.), Ethnocultural aspects of
posttraumatic stress disorder: Issues, research, and clinical applications
(pp. 439-457). Washington, DC: American Psychological Associa-
tion.
SEIDEL, R.W., GUSMAN, F.D., & ABUEG, F.R. (1994). Theoreti-
cal and practical foundations of an inpatient post-traumatic stress
disorder and alcoholism treatment program. Psychotherapy, 31, 67-
78.
WOODWARD, S.H., DRESCHER, K.D., MURPHY, R.T., RUZEK,
J.I., FOY, D.W., & ARSENAULT, N.J. (1997). Heart rate during
group flooding therapy for PTSD. Integrative Physiological and
Behavioral Science, 32, 19-30, 75-83.
WOODWARD, S.H., WANG, S., MASON, J.W., & ARSENAULT,
N.J. (1997). Prediction of REM sleep percent in PTSD. Sleep
Research, 26, 306.
ZAIDI, L.Y. & FOY, D.W. (1994). Childhood abuse experiences
and combat-related PTSD. Journal of Traumatic Stress, 7, 33-42.
Prev Page 8 Next
PILOTS UPDATE
With the next update to the PILOTS database, scheduled
for the end of April, we shall be using many new descriptors
to indicate more precisely the subject-matter of the publica-
tions we are indexing. These new descriptors, as well as
changes to existing ones, will be included in a revision of the
PILOTS Thesaurus. A preliminary version of this will appear
on our Web site at the beginning of May. Later this year a new
version of the PILOTS Database User's Guide will be published
in print and on the Web.
Changing a controlled indexing vocabulary is a difficult
process, requiring careful examination of the existing litera-
ture, educated guesses as to its future development, and
time-consuming consultation with experts in specialized
areas.
If it is decided to apply changes retrospectively, the burden
increases. It is necessary to conduct many searches of the
existing database, examine the documents retrieved by these
searches, decide in which cases existing indexing must be
modified, and then enter the changes in the affected records.
This is so laborious a process that many databases do not
apply changes to their indexing vocabularies to existing
records. Users of the PsycINFO database have to consult the
Thesaurus of Psychological Indexing Terms to learn, for ex-
ample, that they must "use TRAUMATIC NEUROSIS or STRESS
REACTIONS to access references from 1973-1984" because the
descriptor POSTTRAUMATIC STRESS DISORDER was not added to the
Thesaurus until 1985.
Other databases, most notably MEDLINE, revise their
indexing vocabularies frequently and then apply these
changes to all records. Each year when the new edition of
Medical Subject Headings is published, changes are made
throughout the database to reflect the new MeSH terms. In
order to make searching the traumatic stress literature as
easy as possible for our users, we are following this model in
the PILOTS database.
What sort of changes are we making in our indexing
vocabulary?
Many of the new terms will make it easier to find publica-
tions on specific populations affected by traumatic events.
These include ethnic groups such as AMERASIANS, HMONG, and
MIEN; occupational groups such as AGRICULTURAL WORKERS,
JOURNALISTS, and RELIEF WORKERS; religious groups such as
BUDDHISTS, MUSLIMS, ROMAN CATHOLICS, and SIKHS; and other
categories such as GRANDCHILDREN, RUNAWAYS, and RURAL POPU-
LATIONS.
We are adding several new descriptors for traumatic events,
including DENTAL PROCEDURES, GENITAL MUTILATION, JURY SERVICE,
STALKING, and WORKPLACE VIOLENCE. To enhance retrieval of
publications on the effects of exposure to such events, we are
adding ATTACHMENT BEHAVIOR, GASTROINTESTINAL SYMPTOMS, MUL-
TIPLE CHEMICAL SENSITIVITY, and SHAME. And we shall be using
the descriptors TRAUMATIC NEUROSES and WAR NEUROSES to de-
scribe the content of literature from before the DSM era. In the
areas of assessment and treatment, we are adding CROSS
CULTURAL ASSESSMENT, DRAMA THERAPY, MANUAL-BASED TREATMENTS,
THOUGHT FIELD THERAPY, TWELVE STEP PROGRAMS, and others.
We are also changing some existing descriptors to make
them more useful in searching. Publications previously in-
dexed under PSYCHOANALYSIS will instead be assigned PSYCHO-
ANALYTIC THEORY or PSYCHOANALYTIC PSYCHOTHERAPY. We shall
use the new descriptor DISSOCIATIVE SYMPTOMS to describe
studies reporting sequelae of traumatic events that do not
meet DSM criteria for the full-blown disorder, leaving the
existing descriptor DISSOCIATIVE DISORDERS to be used in a more
narrow sense.
These are only a few of the changes that we shall be making
in our indexing vocabulary, and it is possible that some of
these will appear in different form in the new PILOTS The-
saurus. Our hope-and our reason for undertaking this
laborious effort-is that these changes will make it easier to
use the PILOTS database to find the information needed by
researchers, clinicians, and others concerned with PTSD and
other disorders associated with exposure to traumatic events.
Please remember that these changes will take effect only
with the April update to the database. Until the end of April,
continue to use the existing PILOTS Thesaurus as contained
in the Pilots Database User's Guide, Second Edition, November
1994.
|
