International Journal of Law and Psychiatry, Vol. 20, No. 4, pp. 445–458, 1997
Copyright © 1997 Elsevier Science Ltd
Printed in the USA. All rights reserved
Correlates of Perceived Coercion During
Psychiatric Hospital Admission
Michele Cascardi* and Norman G. Poythress**
The use of coercion in the treatment of individuals with severe and persistent
mental illness has been one of the most controversial issues in psychiatry and
mental health law in the United States. Historically, medical determinations of
the need for treatment sufficed as reason to force an individual to receive
mental health care. The American civil rights movement of the 1960s coupled
with exposure of dismal conditions in U.S. state psychiatric hospitals shifted
decisions about commitment from the medical community to the legal com-
munity. While the legal model was intended to prevent unnecessary coerced
treatment and afford the dignity of due process, mental health professionals
have been largely opposed to this model. Such opposition is predicted on the
belief that the adversary process forces the mental health professional and
patient into an oppositional relationship that may enhance a patient’s feeling of
coercion and destroy or damage the therapeutic alliance, resulting in poorer treat-
ment outcomes for the patient than if the adversarial process had been avoided
(Group for the Advancement of Psychiatry, 1994). More recently, individuals in
the consumer movement have further underscored the potentially damaging
effects of coerced treatment and strongly articulated the need for patient partici-
pation and voice in treatment decisions (Campbell, 1992; Leete, 1988).
Until recently, however, limited attention has been directed to the empirical
study of coercion. Studies examining the effects of coercion on patients’ attitudes
toward hospitalization or treatment compliance typically equate coercion with
involuntary legal status (Beck & Golowka, 1988; Kane et al. 1983). However,
there are several limitations in presuming equivalence between coercion and in-
voluntary status. A patient’s legal status frequently changes over time (Cuffel,
*President, Dating Violence Prevention Project, Bala Cynwyd, PA, USA. **Professor, Department of
Mental Health Law and Policy, University of South Florida, Tampa, FL, USA.
Portions of this paper were presented at the 103rd Annual Convention of the American Psychological
Association, New York, August 1995.
Address correspondence and reprint requests to Michele Cascardi, Dating Violence Prevention Project,
544 Conshohocken State Road, Bala Cynwyd, PA 19004 USA.
446M. CASCARDI and N. G. POYTHRESS
1992); many patients initially detained for an involuntary evaluation may be
“pressured” by hospital staff to assent to remaining in the hospital for treat-
ment (Reed & Lewis, 1990). Further, coercion defined objectively in terms of
formal legal status does not always comport with patients’ subjective experi-
ence; some studies have shown that many patients are unaware of their legal
status (e.g., Monahan et al., 1995a), while other studies have found that a sig-
nificant proportion of committed individuals did not feel coerced, whereas a
substantial minority of legally voluntary patients did (e.g., Rogers, 1993).
Limitations such as these have led to systematic examination of the patient’s
experience of coercion associated with the process of hospital admission
(Hoge et al., 1993; Lidz et al., 1993). Research conducted as a part of the Mac-
Arthur Foundation Research Network on Mental Health and the Law estab-
lished measures of coercion using “focus groups” of patients, family members,
and mental health professionals. Key features of coercion include the amount
of influence, control, choice, and freedom an individual perceives to have had
during admission (Gardner et al., 1993).
In addition to measuring the subjective experience of coercion, the Mac-
Arthur Research Network identified factors in the community and hospital
that relate to whether one perceives hospitalization as coercive. Among the
factors are how fairly and respectfully one is treated, the degree to which one
has a “voice” in treatment decisions, as well as the amount of force, threats,
and persuasion used by individuals associated with hospital admission (Mona-
han et al., 1995b). These factors are supported by procedural justice theory
and are strongly related to calls by consumers for participation, voice, and dig-
nity in the treatment process. Monahan et al. (1995b) also suggest that person-
ality characteristics such as external attributional style may influence coercion.
That is, individuals who perceive themselves as being controlled by powerful
others, or as having limited control over things that happen, may be inclined
toward feeling coerced, oppressed, and lacking freedom, irrespective of the
specific characteristics of their situation.
To date, two studies have examined correlates of perceived coercion in a
sample of 157 newly admitted patients either to a state psychiatric hospital or
urban psychiatric treatment facility in the United States (Lidz et al., 1995;
Monahan et al., 1995a). Two groups of patients were examined, those volun-
tarily seeking treatment and those brought to the hospital for involuntary
treatment; all patients were interviewed within 24 hours after their admission.
Among voluntary and involuntary patients, those who perceived themselves
to have been treated fairly, respectfully, and with dignity (i.e., with more pro-
cedural justice) were less likely to feel coerced during admission; however, this
association was far stronger for involuntary than voluntary patients. In addi-
tion, for involuntary patients, those who experienced use of threats and force
were more likely to feel as though they had been coerced during admission
(Lidz et al., 1995). Monahan et al. (1995a) examined the relationship between
pressures and perceived coercion in more depth. Their initial findings sug-
gested that “positive” pressures (e.g., persuasion and inducements) may offset
the coercive effects of “negative” pressure (e.g., threats and force). The current
study was designed to replicate and expand upon the work by the MacArthur
The MacArthur studies included a static assessment of legal status; study
subjects were designated as either voluntary or involuntary based on their le-
gal status at the time of admission. Data from the counties in the state of Flor-
ida in which we have been working (as well as data from prior studies noted
above) clearly indicate that a majority of persons detained for an involuntary
evaluation are permitted to sign into the hospital as voluntary within 1 or 2
days postadmission. Given the dynamic characteristic of legal status, we were
interested in evaluating patients who were ordered by the court to undergo an
the initial “emergency” detainment expired. Of
particular interest was whether individuals who were permitted to sign them-
selves into the hospital as voluntary patients experienced the hospital admis-
sion process differently from those for whom petitions for involuntary treat-
ment were initiated. This comparison is important as many have suggested
that “coerced” voluntary admission may be as, or more, detrimental than “co-
erced” involuntary admission.
A second objective was to evaluate whether patients’ experiences with per-
sons from the community differed from those with hospital staff in predicting
coercion. The quality of experience (procedural justice) during admission sig-
nificantly predicted perceived coercion in previous work (Lidz et al., 1995);
however, in the MacArthur studies, information was averaged across all indi-
viduals with whom the patient had contact during admission. It is possible that
perceptions of treatment by those in the community (e.g., police, family mem-
bers) may differ in important ways from perceived treatment by those in the
hospital (e.g., psychiatrist, nurse, social worker). Two global measures of
the process and pressures experienced during admission were examined in the
current study. One version included questions regarding an individual’s inter-
actions with people in the community like family, friends, police; the other so-
licited perceptions of treatment by hospital staff like psychiatrists, nurses, so-
cial workers, etc. Finally, we evaluated the contributions of locus of control
and psychiatric symptom severity to perceptions of coercion to determine
whether such perceptions were a function of personality or clinical symptoms
(e.g., psychosis or paranoia) rather than situation-specific characteristics of
Participants were recruited from the crisis stabilization unit (CSU) at two
community mental health centers in west central Florida. The CSU serves as a
receiving facility for Florida’s Department of Health and Rehabilitation Ser-
vices (HRS). Receiving facilities are those agencies designated by HRS to
conduct psychiatric evaluations within 72 hours of admission for persons ad-
mitted via court order to undergo an involuntary psychiatric evaluation. All
participants were individuals ordered to CSUs by the court for involuntary
evaluation and were interviewed soon after admission (mean
1.86 days). Approximately half the participants (58 and 62) were re-
cruited from each CSU, between October 1994 and June 1995. Half of the pa-
448M. CASCARDI and N. G. POYTHRESS
tients at each site were judged by the psychiatrist (upon completion of the 72-
hour involuntary evaluation) to require involuntary commitment via the court
(involuntary), and the other half were permitted to sign into the CSU as vol-
untary patients (voluntary). All participants were 16 years or older.
The clinical and demographic characteristics of the sample are presented in
Table 1. The involuntary and voluntary groups were highly similar on demo-
graphic measures; the majority of persons were Caucasian, in their mid-thirties,
and male. About 75% of those in each group were single or divorced, and
about two-thirds of each group were unemployed prior to the current admis-
sion. The involuntary group appeared, based on self-report data, to have more
education than the voluntary group. Clinically, the groups were similar in
Mean age (years)
% Unemployed prior to admission
% Receiving outpatient treatment
prior to admission
BPRS-A mean scores
PERCEIVED COERCION 449
terms of chart diagnosis and in involvement in outpatient treatment prior to
admission. Voluntary patients had slightly lower scores on all psychiatric
symptom indices except for the depression subscale.
MacArthur Admission Experience Interview (AEI).
et al., 1993, for exact interview items) consists of three sections: (1) treatment
by others during admission process (“procedural justice”), (2) pressures used
during admission, and (3) perceived coercion. In previous research with the
AEI, respondents were queried separately about each individual associated
with their admission. Then, responses to each item were averaged across indi-
viduals. We modified the AEI protocol such that inquiries about perceptions
of procedural justice and pressures used during admission were directed to-
ward two sets of individuals: (1) all persons from the community with whom
the patient came into contact during admission and (2) all hospital staff with
whom the patient came into contact. Participants were also asked to identify
whether a particular person from the community and/or a particular person
from the hospital staff most influenced their responses. If they replied in the
affirmative, participants identified who it was.
The procedural justice instruments consist of six items related to procedural
justice constructs (motivation, respect, voice, validation, fairness, freedom),
five of which are rated on a 4-point scale from “never/not at all/none” to “very
much/very.” For example, motivation is measured by, “To what extent did
persons from the community (or hospital staff) who were involved in your cur-
rent admission do what they did out of concern for you?” Choices ranged
from: “very much; mostly; a little; not at all.” Freedom, the sixth item, assessed
patients’ perceptions of whether they would be released from the hospital if
they so desired on a 4-point scale from “definitely yes” to “definitely not”;
This item was not used by Lidz et al. (1995). The last item on the proce-
dural justice scale asked whether the patient felt as if he or she had been de-
ceived during admission. Responses to the seven items were summed to create
an unweighted linear composite of procedural justice; higher scores indicate
more perceived procedural justice. This method differed from Lidz et al.
(1995); they used principal components analysis to create a weighted summa-
Participants then responded to four items regarding whether any of four
types of pressures had been used to convince them to enter the hospital: per-
suasion, promises, threats, and force. Using the rationale of Monahan et al.
(1995a), pressures were classified as “positive” (persuasion and promises) and
“negative” (threats and force). For purposes of data analysis, positive and neg-
ative pressures were dummy-coded as present or absent.
Perceived coercion is a 4-item measure that assesses the degree to which the
patient feels free to refuse psychiatric hospital admission (e.g., “How free
were you to do what you wanted about coming into the hospital?”). Respon-
dents rate each item using a 4-point scale from “not al all” to “very.” Corre-
spondence analysis indicated that the items are internally consistent and re-
flect one latent dimension (Gardner et al., 1993).
The AEI (see Gardner
450 M. CASCARDI and N. G. POYTHRESS
Each measure of the Admission Experience Interview—perceived proce-
dural justice, pressures, and perceived coercion—has demonstrated accept-
able levels of stability when assessed within a 24–48-hour time frame. Stability
coefficients range from .62 to .72 (Cascardi, Poythress, & Ritterband, in press).
Locus of Control (LOC).
report measure in which individuals rate the strength of their beliefs on a 7-point
Likert scale (“strongly agree” to “strongly disagree”). This abbreviated LOC
scale was developed as an alternative to Levenson’s (1974) 24-item LOC scale.
The factor structure and psychometric properties of Levenson’s LOC scale
have been well documented (e.g., Blau, 1984). Sapp and Harrod (1993) repli-
cated the factor structure of the longer LOC scale and reported adequate reli-
ability and construct validity of their 9-item LOC scale.
LOC (Sapp & Harrod, 1993) is a 9-item self-
Brief Psychiatric Rating Scale-Anchored Version (BPRS-A).
a modified version of the BPRS (Overall & Gorham, 1962), which is the most
widely used quantitative scale for rapid clinical assessment of inpatients’ men-
tal state (Hafkenscheid, 1991). The scale consists of 18 symptom constructs.
Each construct is rated on a scale of increasing severity from one to seven. The
instrument is completed following a 15–20-minute semi-structured interview.
The BPRS-A, designed to increase inter-rater reliability, includes anchors for
each scale point, a guide of symptom probes, and expanded and elaborated
symptom descriptions (Woerner, Mannuzza, & Kane, 1988). A total psychiat-
ric symptom severity score as well as four subscales scores—psychoticism, de-
pression, hostility, and emotional withdrawal—are computed (Overall & Por-
terfield, 1963). Inter-rater agreement in the current study, as measured by
perfect agreement on the Lawlis and Lu T statistic (Tinsley & Weiss, 1975),
ranged from .22 to .84 (mode
.74). When the criterion for inter-rater agree-
ment was relaxed so that agreement was considered to be within one scale
point, T values ranged from .64 to 1.00 (mode
coefficient for BPRS total score was .96.
The BPRS-A is
1.00). Intraclass correlation
Recruitment of involuntary subjects involved monitoring the dispositional
decisions made by psychiatrists at the 72-hour evaluation through frequent
telephone contact with social work staff and CSU discharge coordinators. Un-
less staff advised that the patient was too aggressive or unstable to participate,
all eligible patients were asked to participate. Recruitment procedures for vol-
untary patients were similar to those used with involuntary patients; however,
not all voluntary patients were similar to those used with involuntary patients;
however, not all voluntary patients were recruited into the study. Because the
design required equal numbers of involuntary and voluntary patients, we re-
cruited voluntary patients (of whom there was a much larger number) at ap-
proximately the same pace as involuntary patients. Overall, approximately 30%
of the patients approached declined participation. Although this loss of poten-
tial subjects is higher than is desirable, our refusal rate is comparable to rates re-
ported in previous studies by Applebaum, Mirkin, and Bateman (1981) (33%)
and Levine et al. (1994) (31%). Fifteen potential participants (4 voluntary and 11
involuntary) were excluded as they were too aggressive or unstable to participate.
After a patient was successfully recruited into the study, and written in-
formed consent was obtained, demographic and social history information was
recorded via patient self-report and chart review and the research protocol
was administered. All protocol items were read aloud to the patient while he
or she followed along in a participant’s manual.
Involuntary vs. Voluntary
One-way multivariate analysis of variance was used to evaluate group dif-
ferences on admission experience variables. There was a significant overall
group effect, Wilks’s lambda (7, 112)
cated that involuntary subjects experienced admission as significantly more
negative, feeling as though they were treated with less respect, fairness, care,
and taken less seriously than did voluntary subjects in interactions in both the
community and the hospital,
.01, respectively. Involuntary subjects also reported significantly more
threats and force than did voluntary subjects,
.05, respectively. Finally, involuntary subjects felt signifi-
cantly more coerced,
Mean group differences are presented in Table 2.
.003. Post hoc analyses indi-
.001, than did voluntary subjects.
Zero-order Pearson Product-Moment correlations
factors associated with perceived coercion. Perceived coercion was signifi-
were used to evaluate
Involuntary vs. Voluntary: Means and Standard Deviations
Positive pressures in community
Negative pressures in community**
Positive pressures in hospital
Negative pressures in hospital***
Point biserial correlations were used when the association between a dichotomous variable (e.g., sex,
race) and perceived coercion was computed. Positive and negative pressures were also treated as dichoto-
mous variables (presence/absence) in the correlational and multiple regression analyses.
452 M. CASCARDI and N. G. POYTHRESS
cantly associated with community and hospital process (
community and hospital negative pressures (
.35), and education (
with site (
.06), age (
.13), sex (
.02). The correlation matrix is presented in Table 3.
.37), legal status
.22). coercion was not significantly associated
.09), diagnosis (
.03), or locus of
Hierarchical regression analyses were used to evaluate incremental contri-
butions of demographic, clinical, and various aspects of hospital admission to
the prediction of perceived coercion. Locus of control and psychiatric diagno-
sis were not included in the regression analyses given their lack of bivariate as-
sociation with perceived coercion. “Nuisance” variables (i.e., demographic
and clinical variables) were added to the model first.
Demographics (age, race [white/nonwhite], sex, and educational level) were
entered in the first block. These variables accounted for 6% of variance and
did not significantly improve prediction of perceived coercion. In the second
step, psychiatric symptom severity was added to the model, but it also did not
significantly improve prediction of perceived coercion. In the third step, legal
status was added to the model. Legal status significantly improved prediction
of coercion, increasing the percent of variance explained by almost 10%.
Based on earlier work by Lidz et al. (1995), the hospital and community proce-
dural justice measures were added next to the regression equation followed by
Correlates of Perceived Coercion
1 Perceived coercion
2 Process community
3 Positive community
4 Negative community
5 Process hospital
6 Positive hospital
7 Negative hospital
8 Legal status
10 Locus of control
?0.0130.370* 0.553* 0.319*
?0.283* ?0.247* ?0.038
*p ? .01.
**p ? .05.
hospital and community pressures.2 On the fourth and fifth steps, hospital and
community procedural justice measures were added. Each procedural justice
variable significantly improved prediction of perceived coercion, and yielded a
22% net increase in variance explained. On the last two steps, hospital and
community pressures were added. Hospital pressures did not significantly im-
prove prediction of coercion.
In contrast, negative community pressures (threats and force) did signifi-
cantly improve the prediction of perceived coercion, increasing the percent of
variance explained by 9%. In the final model almost 49% of variance (43%
adjusted variance) was explained. Table 4 illustrated R2, increment in R2, B at
initial entry to the regression equation, and B in the final model. As indicated
in the table, the only variables that significantly predict perceived coercion in
the final model are community process (B ? .34, p ? .001) and negative com-
munity pressures (B ? .38, p ? .001).
To contextualize the above findings, exploratory analyses were conducted
to determine whether mention of particular classes of individuals from the
Sex (0, 1)
Race (0, 1)
F (12, 107) ? 8.50; Adjusted R2 ? .431; Multiple R ? .699.
*p ? .0001.
2Hospital process was added to the regression equation first because it shares almost all of its variance
with community process. By entering hospital process before community process, the contribution of hospi-
tal process is preserved.
454 M. CASCARDI and N. G. POYTHRESS
community and hospital produced different patterns of association of proce-
dural justice and pressures with perceived coercion. Regarding interactions
with members of the community, 8.3% (n ? 10) of subjects mentioned the po-
lice, 35% (n ? 42) family/friends, and 11.7% (n ? 14) mentioned community
providers (e.g., counselor, case manager) as having the most influence on their
perceptions of procedural justice; 43.3% (n ? 52) did not single out any one
particular individual. Although too few subjects spontaneously mentioned a
police officer or community provider for detection of reliable associations,
some interesting trends were noted in the data (correlations with perceived
coercion are presented in Table 5). A very strong association existed between
negative pressures and perceived coercion when a police officer or community
provider was mentioned (r ? .78 and r ? .80, respectively). When a family/friend
was mentioned, a significant association occurred between positive pressures
and perceived coercion such that the more persuasion or promises offered, the
less coerced the patient felt (r ? ?.34, p ? .05). This association was small when
police officers, community providers, or “no one” was mentioned by the patient.
Regarding experiences with hospital staff, 11.7% of patients mentioned
their psychiatrist (n ? 14), 25.8% mentioned a nurse/mental health technician/
counselor/social worker (n ? 31), and 62.5% (n ? 75) did not mention any-
one. Too few patients mentioned the psychiatrist to detect reliable associa-
tions with perceived coercion; however, as above, some interesting trends
were noted. Of particular interest was that the more persuasion or promises
used by the psychiatrist the more coerced the patient felt (r ? .42, ns).
There were three primary objectives of the present study: replication of Lidz
et al. (1995); examination of differences in perceptions about hospital admission
for voluntary and involuntary patients, all of whom had been involuntarily de-
tained initially for a psychiatric evaluation; and examination of the impact of two
broad classes of individuals (community members (e.g., family, police) and hos-
(n ? 10)
(n ? 42)
(n ? 14)
(n ? 52)
(n ? 14)
(n ? 31)
(n ? 75)
*p ? .01.
**p ? .05.
F/F ? Family/Friends
pital staff (e.g., psychiatrist, nurses), locus of control, and psychiatric symptom se-
verity on patients’ perceptions of coercion. Overall, our results substantiate ear-
lier findings of Lidz et al. (1995) in that legal status (involuntary), procedural
justice, and negative pressures are significantly related to perceived coercion.
In regard to our second objective, although all patients in the present study
had been detained involuntarily initially, those permitted to sign as voluntary
perceived their hospital admission as less coercive and more respectful, fair,
and dignified than did those against whom petitions for involuntary treatment
had been filed. “Voluntary” patients were also less likely to have perceived
themselves as threatened or forced into the hospital by community members
or hospital staff. Although we cannot evaluate whether persons who assent to
voluntary status perceive hospitalization as less coercive than do those volun-
tarily seeking treatment in the absence of any legal coercion, it is clear that the
experiences of patients converted to voluntary status are less negative than
those against whom petitions for involuntary commitment are filed. Anecdot-
ally, patients permitted to sign as voluntary are those who are “more coopera-
tive” with hospital procedures, i.e., they are more willing to take medication
(Poythress, Cascardi, & Ritterband, 1996).
In contrast, those who persist in their refusal to sign as voluntary or cooper-
ate with treatment often have petitions for involuntary treatment filed against
them. Though we cannot assert that voluntary patients signed as voluntary be-
cause they perceived themselves as less coerced and treated with more proce-
dural justice, it is plausible that such treatment may lead to greater receptivity
to psychiatric hospitalization and thus greater likelihood of conversion to vol-
untary status. Alternatively, it is plausible that voluntary patients re-evaluated
their admission experience in light of their change in legal status. That is, as
cognitive dissonance theory would suggest, patients who elected to submit to
voluntary hospitalization may have subsequently reframed their admission ex-
perience as more positive. This is also somewhat consistent with findings re-
garding Stone’s “Thank-You” Theory in which patients whose symptoms were
substantially reduced at the conclusion of hospitalization maintained more
positive views of their inpatient treatment than did those who failed to achieve
similar gains (Beck & Golowka, 1988).
To understand factors associated with coercion in more detail than in previ-
ous work (e.g., Lidz et al., 1995), personality and symptom characteristics were
added to the list of possible predictors of coercion, and community and hospital
experiences were evaluated separately. Locus of control and symptom severity
were unrelated to perceived coercion; thus, coercion does not appear to reflect
a general inclination toward feeling controlled by others, nor does it appear to
bear an association to severe psychosis, paranoia, or depression, symptoms that
may enhance one’s feeling of powerlessness. Examination of the perceived in-
fluence of community members and hospital staff suggest that procedural jus-
tice and negative pressures from community members bore a stronger associa-
tion with perceived coercion than did experiences with hospital staff involved
in the patient’s admission. Development of strategies to decrease perceived co-
ercion may then imply diverse intervention strategies depending upon the per-
sons with whom the patient has had contact during admission. The finding that
experiences with community members were more strongly associated with per-
456M. CASCARDI and N. G. POYTHRESS
ceived coercion is particularly interesting in that much attention has been di-
rected toward treating clinicians’ behavior and coercive psychiatric treatment.
Because it was somewhat surprising to find that hospital staffs’ behavior was
less strongly associated with perceptions of coercion than behavior of commu-
nity members associated with the patients’ admission, exploratory analyses
were conducted to shed additional light on these associations.
Exploratory analyses suggested that the pattern of association of positive
pressures with perceived coercion depended upon who substantially influ-
enced patients’ responses on the admission experience questionnaires. Specif-
ically, for patients who spontaneously stated that family or friends most influ-
enced their responses to protocol items, persuasion and promises were
associated with reduced perceptions of coercion. When these same behaviors
were emitted by a psychiatrist, the resulting association with perceived coer-
cion was in the opposite direction; persuasion and promises were associated
with heightened perceptions of coercion. The above patterns must be inter-
preted with caution as the magnitude of the associations did not reach statisti-
cal significance, most likely owing to the small sample size. Increased negative
pressures were consistently associated with increased feelings of coercion irre-
spective of the persons encountered during the admission experience; how-
ever, this association was particularly strong when police or community-based
providers were spontaneously mentioned.
One significant limitation of these analyses is that we do not know why cer-
tain patients mentioned someone in particular while others did not; nor do we
know the quality of patients’ ongoing relationship with persons mentioned. It
is possible that those with very good or very bad experiences with particular
individuals were mst likely to mention someone in particular. For example,
some patients recalled positive and supportive relationships with mental
health technicians or nurses associated with their current admission. There-
fore, it is difficult to tease apart, in a systematic manner, different patterns of
association based on the person with whom the patient interacted owing to
limited sample size and method of inquiry.
The instruments developed by the MacArthur Research Network on Men-
tal Health and the Law provide researchers with standardized measures for
the systematic and quantified assessment of the experience of procedural jus-
tice characteristics (e.g., respect, voice, fairness) in the interpersonal interac-
tions associated with perceived coercion in the process of psychiatric hospital-
ization. The studies reviewed here and the data from the present study suggest
several directions for future research related to perceived coercion.
One obvious direction is for researchers to observe the objective behaviors
of persons in the community or hospital staff involved in implementing hospi-
tal admission to determine whether and to what degree patients’ subjective ex-
perience of being coerced or pressured into treatment corresponds to, or is the
outgrowth of, identifiable actions by persons in the community or hospital
staff. If reliable methods for rating and coding such behaviors can be devel-
oped, then those experienced by patients as coercive can be targeted for inter-
vention with community persons and providers in an effort to reduce the coer-
cive quality of the admissions process. It may be more difficult to do
prospective studies of interactions with key persons in the community, for of-
ten “cases” for research evaluation are identified only after events of some sig-
nificance in the community trigger attempts at hospitalization (e.g., emergency
room visits, petitions for involuntary evaluation or commitment); this is prob-
lematic in that our data suggest that community process and negative commu-
nity pressures are most strongly associated with perceived coercion. Prospec-
tive studies of hospital staff behavior, however, should be possible, and
reconstructions of interactions with people in the community (family mem-
bers, police, community care-givers) may provide some clues to specific be-
haviors that individuals with mental illness experience as coercive.
A second need is for research that examines the association between per-
ceived coercion and treatment outcome and participation. Empirical examina-
tion of the popular belief that patients who subjectively experience a high de-
gree of coercion may be less likely to participate in treatment and perhaps
more likely to suffer worse treatment outcomes is needed. Various measures
of treatment participation (e.g., medication refusal, attendance at/participa-
tion in assigned therapies, etc.) and treatment outcome (e.g., length of stay,
symptom reduction, satisfaction with services) are available as dependent
measures for those who may investigate either the short-term or long-term ef-
fects of perceived coercion.
Methods proposed to reduce the experience of coercion in the admissions
and treatment process should also be investigated. Following Tyler’s theoreti-
cal analysis of procedural justice in the context of civil commitment (Tyler,
1992), we are currently examining patients’ perceptions of involuntary com-
mitment hearings, using videotaped vignettes of mock hearings in which we
manipulate the procedural justice characteristics of commitment hearings
(e.g., enhanced/reduced “voice” as reflected in the quality of attorney repre-
sentation; enhanced/reduced “respectful treatment” by the hearing officer),
and the degree of perceived coercion associated with these vignettes.
Other researchers (Mazade, Blanch, & Petrila, 1994) have suggested that
the process of mediation might be attempted at a number of different points in
the treatment process (e.g., to resolve disputes in the community with land-
lords or case managers; at negotiations about whether to voluntarily enter an
inpatient unit; in treatment planning and implementation within a residential
program, such as the application of seclusion or restraints) with the objective
of enhancing patients’ sense of empowerment, participation and self-determi-
nation, and reduction in perceived coercion.
Mediation (and other alternative dispute-resolution procedures) offers po-
tentially new ways for mental health professionals, patients (“consumers” of
mental health services), and community providers and support persons to in-
teract that break the traditional roles mandated by the customary clinical and
legal mechanisms for making decisions about whether and how care will be
provided. Whether such novel ways of interacting will reduce perceived coer-
cion, however, remains subject to empirical investigation.
Applebaum, P. S., Mirkin, S. A., & Bateman, A. L. (1981). Empirical assessment of competency to consent
to psychiatric hospitalization. American Journal of Psychiatry, 138, 1170–1176.
458 M. CASCARDI and N. G. POYTHRESS Download full-text
Beck, J. C., & Golowka, E. A. (1988). A study of enforced treatment in relation to Stone’s “thank-you” the-
ory. Behavioral Sciences and the Law, 6, 559–566.
Blau, G. J. (1984). Brief note comparing the Rotter and Levenson measure of locus of control. Perceptual
and Motor Skills, 58, 173–174.
Campbell, J. (1992). The Well-Being Project: Mental health clients speak for themselves. National Associa-
tion of State Mental Health Program Directors Research Institute Conference Proceedings, 31.
Cascardi, M., Poythress, N. G., & Ritterband, L. M. (in press). Stability of psychiatric patients’ perception of
coercion. Journal of Clinical Psychology.
Cuffel, B. (1992). Characteristics associated with legal status change among psychiatric patients. Community
Mental Health Journal, 28, 471–482.
Gardner, W., Hoge, S. K., Bennett, N. S., Roth, L. H., Lidz, C. W., Monahan, J., & Mulvey, E. P. (1993).
Two scales for measuring patients’ perceptions of coercion during mental hospital admission. Behavioral
Sciences and the Law, 11, 307–321.
Group for the Advancement of Psychiatry. (1994). Forced into treatment: The role of coercion in clinical
practice. Washington, DC: American Psychiatric Press.
Hafkenscheid, A. (1991). Psychometric evaluation of a standardized and expanded Brief Psychiatric Rating
Scale. Acta Psychiatrica Scandanavia, 84, 294–300.
Hoge, S. K., Lidz, C. W., Mulvey, E. P., Roth, L. H., Bennett, N. S., Siminoff, A., Arnold, R., & Monahan, J.
(1993). Patient, family, and staff perceptions of coercion in mental hospital admission: An exploratory
study. Behavioral Sciences and the Law, 11, 281–293.
Kane, J. M., Quitkin, F., Rifkin, A., Wegner, Rosenberg, G., & Borenstein, C. (1983). Attitudinal changes of
involuntarily committed patients following treatment. Archives of General Psychiatry, 40, 373–377.
Leete, E. (1988). The role of the consumer movement and persons with mental illness. Switzer Seminar 13, 8-32.
Levenson, H. (1974). Activism and powerful others: Distinctions with the concept of internal-external con-
trol. Journal of Personality Assessment, 38, 377–383.
Levine, S., Byrne, K., Wilets, I., Fraser, M., Leal, D., & Kato, K. (1994). Competency of geropsychiatric
patients to consent to voluntary hospitalization. The American Journal of Geriatric Psychiatry, 2, 300–308.
Lidz, C. W., Mulvey, E. P., Arnold, R., Bennett, N. S., & Kirsch, B. (1993). Coercive interactions in a psychi-
atric emergency room. Behavioral Sciences and the Law, 11, 269–280.
Lidz, C. W., Hoge, S. K., Gardner, W. P., Bennett, N. S., Monahan, J., Mulvey, E. P., & Roth, L. H. (1995).
Perceived coercion in mental hospital admission: Pressures and process. Archives of General Psychiatry,
Mazade, N., Blanch, A., & Petrila, J. (1994). Mediation as a new technique for resolving disputes in the
mental health system. Administration and Policy in Mental Health, 21, 431–445.
Monahan, J., Hoge, S. K., Lidz, C. W., Eisenberg, M. M., Bennett, N. S., Gardner, W. P., Mulvey, E. P., &
Roth, L. H. (1995a). Coercion to inpatient treatment: Initial results and implications for assertive treat-
ment in the community. In D. Dennis & J. Monahan (Eds.), Coercive and aggressive community treat-
ment: A new frontier for mental health law. New York: Plenum.
Monahan, J., Hoge, S. K., Lidz, C. W., Roth, L. H., Bennett, N. S., Gardner, W. P., & Mulvey, E. P. (1995b).
Coercion and commitment: Understanding involuntary mental hospital admission. International Journal
of Law and Psychiatry, 18, 249–263.
Overall, J. E., & Gorham D. R. (1962). The Brief Psychiatric Rating Scale. Psychological Reports, 12, 578–602.
Overall, J. E., & Porterfield, J. L. (1963). Powered vector method of factor analysis. Psychometrika, 28, 415–422.
Poythress, N. G., Cascardi, M., & Ritterband, L. M. (1996). Psychiatric patients’ capacity to consent to vol-
untary hospitalization: A study of psychiatrists’ judgments. Bulletin of the American Academy of Psychia-
try and Law.
Reed, S. C., & Lewis, D. A. (1990, Spring/Summer). The negotiation of voluntary admission in Chicago’s
state mental hospitals. The Journal of Psychiatry and Law, 137–163.
Rogers, A. (1993). Coercion and “voluntary” admission: An examination of psychiatric patient views.
Behavioral Sciences and the Law, 11, 259–267.
Sapp, S. G., & Harrod, W. J. (1993). Reliability and validity of a brief version of Levenson’s Locus of Con-
trol Scale. Psychological Reports, 72, 539–550.
Tinsley, H. E. A., & Weiss, D. I. (1975). Interrater reliability and agreement of subjective judgments. Jour-
nal of Counseling Psychology, 22, 358–376.
Tyler, T. R. (1992). The psychological consequences of judicial procedures: Implications for civil commit-
ment hearings. SMU Law Review, 46, 421–433.
Woerner, M. G., Mannuzza, S., & Kane, J. M. (1988). Anchoring the BPRS: An aid to improving reliability.
Psychopharmacological Bulletin, 24, 112–117.