Process Analysis of Judges’ Commitment Decisions: A Preliminary Empirical
Study
Harold Bursztajn, M.D., Thomas G. Gutheil, M.D., Mark Mills, J.D., M.D.,
Robert M. Hamm, Ph.D., and Archie Brodsky, B.A.
Presented at the 137th annual meeting of the
American Psychiatric Association, Los Angeles, May 5-11, 1984. Received
Jan. 31, 1985; revised Aug. 7, 1985; accepted Sept. 16, 1985. From the
Program in Psychiatry and the Law, Massachusetts Mental Health Center;
the Department of Psychiatry, University of California, Los Angeles,
School of Medicine; the Center for Research on Judgment and Policy, University
of Colorado, Boulder; and Hampstead Hospital, Hampstead, N.H. Address
reprint requests to Dr. Bursztajn, Program in Psychiatry and the Law,
Massachusetts Mental Health Center, 74 Fenwood Rd., Boston, MA 02115.
Supported in part by NIMH grant MH-16460.
The authors thank Paul S. Appelbaum, M.D., for critical review and comment;
the Honorable Maurice Richardson for help and inspiration; and Ms. Audrey
Bleakley for help in the preparation of the manuscript.
The current debate over the "police powers" versus parens
patriae rationales for involuntary hospitalization of the mentally
ill underscores the need for empirical study of the process of judicial
decision making in civil commitment and determinations of competence.
The authors report the ratings on 26 descriptive variables made by five
Massachusetts district court judges for 35 patients in civil commitment
hearings. Nearly all of the hearings resulted in commitments. These findings
suggest that psychiatrists may be setting too high a threshold for petitioning
for commitment. Experienced judges appeared to be sensitive to the kinds
of clinical issues that earlier studies have shown to contribute significantly
to the psychiatrist’s decision to petition for commitment.
(Am J Psychiatry 143:170-174, 1986)
Public policy with respect to the civil commitment of the mentally ill
is currently the subject of intense debate. In recent years the dominant
legal trend has been toward a "police powers" or public safety
rationale, according to which individuals diagnosed as mentally ill (alone
among potentially violent individuals) are subjected to a form of preventive
detention when they are also judged to be dangerous [1].
However, the observed inadequacies of this approach have led to a revival
of interest in the older parens patriae model, according to
which the state is empowered to hospitalize involuntarily, for their
own benefit, individuals in need of treatment. Most clinicians would
prefer to restore need for treatment as a primary justification for involuntary
hospitalization of the mentally ill, leaving the question of dangerousness
in the hands of the criminal justice system [2]. This
preference is reflected in the American Psychiatric Association’s model
state statute on civil commitment [3], which links a
patient’s committability to his or her competence to give informed consent.
In the context of this debate it would be especially valuable to learn
more about how judges actually make commitment decisions and, in particular,
how the question of competence figures in judges’ thinking about committability.
Do judges follow a "police powers" model, a parens patriae model,
or a combination of both? It is, at present, difficult to answer this
question because the relationship between psychiatry and the law has
to date had little illumination in the form of empirical studies of the
process of decision making. Although such research has now begun in psychiatry
[4], the study of judicial decisions has (with the exception
of non-data-based reflections [5]) taken the form of
outcome studies [6], simulation and/or observation studies
[7, 8], and models of process based
on outcome [9]. Yet, studies applying contemporaneous
process measures to actual court hearings would seem essential to understanding
the subtle interplay between clinical and judicial reasoning in areas
such as civil commitment.
In this preliminary empirical study of civil commitment decisions by Massachusetts
district court judges we will attempt a form of in-process examination
that has not been described before. In Massachusetts, a voluntarily hospitalized
patient who has been admitted on a "conditional" basis must
give 3 days’ written notice (a 3-day paper) signifying an intention to
leave the hospital, after which the patient’s psychiatrist has 3 days
to decide whether to petition the court for commitment lasting up to
6 months. Thus, commitment is a two-step process: the physician does
or does not petition; if petitioned, the judge does or does not commit.
Having investigated and reported elsewhere the psychiatrist’s decision
to petition for commitment [4], we sought to find out
whether judges use similar strategies, conscious and unconscious, in
making the actual decision to commit. We expected that our findings might
also shed light on the previously observed tendency of judges to defer
to psychiatrists’ recommendations for commitment, and/or that of psychiatrists
to attempt to anticipate the likely judicial disposition of a case when
deciding whether to petition [8, 10-12].
Psychiatrists’ Decisions On Commitment
In the previous study [4], clinicians were found to be
properly attentive to the three legally mandated criteria [13]
(along with the presence of serious mental illness) for civil commitment
in Massachusetts: dangerousness to self, dangerousness to others, and
inability to care for self. Perceived patient characteristics that were
clearly related to these criteria (e.g., having a place to live, having
support on the outside, able to work, crazy or sane) also strongly influenced
the decision to petition. Further analysis of the study (our unpublished
1983 data) revealed the usefulness of the clinicians’ introspective reports
about the relative importance of the various factors in the decision
to petition. These self-reports gave insight into the clinicians’ decisions
that was not available with the descriptive ratings alone. In particular,
they showed that the psychiatrists took into account subjective and interpersonal
factors which are cited as important predictive variables both in the
clinical literature [14] and in empirical studies of
suicide assessment [15] and other prognostic areas
[16]. For example, the psychiatrists reported that
a patient’s
"likability" influenced them in the direction of being less
likely to petition. This may be a valid consideration, since likability
reflects both the strength of the therapeutic alliance and the patient’s
presumed capacity to inspire loving support outside the hospital. Conversely,
an unlikable patient would be expected to be at higher risk.
These findings demonstrated the value of asking two kinds of questions
(both descriptive and introspective) to capture the various features
of decision making. For this reason, as well as to provide a close comparison
with the data on the psychiatrists’ decisions, we gave the judges in
the present study the same type of questionnaire, one that elicited their
assessments of the patients before them and of the weight they would
assign to each of the factors contributing to their own decisions.
Method
Five Massachusetts district court judges, with an average of 7 years’
experience on the bench, participated in the study as a contribution
to their continuing education. The respondents represented a sample of
the estimated 20 to 30 judges involved in commitments in Massachusetts.
They were asked to fill out an anonymous, four-part questionnaire immediately
after every civil commitment hearing over which they presided during
the 4-month study period. Part 1 consisted of demographic data concerning
the patient, the judge, the petitioning facility, and the hearing. Part
2 (the rating scale) asked the judge to rate the patient on a 7-point
scale for 26 variables (e.g., 1 = frightening, 7 = not frightening; 1
= able to take care of self, 7 = not able to take care of self). Part
3 (the impact scale) asked each judge to go back and assign a weighting
signifying how important each of the 26 factors was in the decision to
commit or not. They were told, "The number -100 means that this
fact about the patient was strongly against commitment in your thinking,
and +100 means that the fact is strongly for committing the patient,
and 0 means the fact has no bearing at all on your decision." Part
4 was a 7-point scale for rating the ease with which the decision was
reached.
The five judges returned 41 questionnaires. Of the 34 that reported a
decision to commit, 32 were filled out sufficiently to be included in
the data analysis. Of the six that reported a continuance of the case,
two were adequately filled out. One questionnaire did not report any
decision but was otherwise complete. Thus, 35 questionnaires were available
for data analysis. With this data base the 26 factors were rank-ordered
according to their reported relative mean impact on the commitment decision,
ranging from those weighing most heavily in favor of commitment to those
weighing most heavily against commitment. We had also intended to determine
the correlation between the patient characteristic scales and the commitment
decision (i.e., commit or release), as had been done in the study of
psychiatrists’ decisions to petition. However, since none of the questionnaires
reported a decision not to commit, such correlations were impossible.
Instead, correlations were calculated between the judges’ ratings of
patient competence and each of the other 25 rating scales, as well as
between competence and diagnosis.
Results
For the total sample of 41 patients the commitment rate was 83%: 34 of
the cases. There were six (15%) continuances and one (2%) unknown. Among
the 35 questionnaires subjected to further analysis, the figures were
32 (91%) committed, two (6%) continued, and one (3%) unknown. Moreover,
in most cases the judges did not find the decision difficult (responses
to part 4 clustered at 1 and 2 on the 7-point scale). On the impact scales,
where the judges rated the relative strength of the factors influencing
them toward commitment, the three factors reported to have the greatest
impact were whether the psychiatrist’s opinion was convincing, whether
the patient would be a reliable outpatient, and whether the patient was
able to take care of himself or herself (table 1). The third is one of
the criteria specified by Massachusetts law. In general, factors having
to do with compliance were more influential than those having to do with
dangerousness, despite the centrality of dangerousness in the legal code.
Finally, the three subjective factors grouped under judge’s reaction
to patient in table 1 had the least impact here, in contrast with the
study of psychiatrists’ decision making [4].
In this sample of 35 patients, all of whom were either committed or held
temporarily pending further disposition, incompetence as perceived by
the judges correlated strongly (p<.01) with three other patient characteristics.
The highest correlation was with unpredictable (r=.53, df=31, p=.001),
followed by not capable of working (r=.47, df=31, p=.003) and not likable
(r=.47, df=26, p=.007). Incompetence did not correlate, however, with
whether the patient was in distress, able to take care of himself or
herself, or any of the factors listed under violence or suicidality.
A further analysis revealed that judgments of competence correlated with
diagnosis, at least to the extent that schizophrenic patients were more
likely to be seen as competent than were those with manic-depression,
borderline personality disorder, or organic illness. The correlations
with incompetence for these diagnostic groups were -.33 (df=30, p=.03),
.25 (df=31, n.s.), -.12 (df=31, n.s.), and .16 (df=31, n.s.), respectively.
TABLE 1. Relationship Between Five Judges' Perceptions of Patient
Incompetence and Patient Characteristics in 35 Civil Commitment Hearings
|
|
Impact on
Judges' Decisions |
Correlation With
Incompentence |
|
|
|
|
|
Characteristic |
Meana |
Rank |
r |
df |
Significance
(one-tailed) |
|
Violence of patient |
|
|
|
|
|
No danger to others |
66.5 |
7 |
-.02 |
30 |
n.s. |
No extra security present |
1.3 |
23 |
-.16 |
31 |
n.s. |
No more than verbal threats |
66.5 |
7 |
.14 |
29 |
n.s. |
No recent violence |
69.1 |
4 |
-.27 |
28 |
n.s. |
Suicidality of patient |
|
|
|
|
|
Not dangerous to self |
57.3 |
10 |
-.14 |
28 |
n.s. |
No self-destructive behavior (only threats) |
50.0 |
11 |
-.14 |
25 |
n.s. |
Self-harm only in remote past (if at all) |
49.5 |
12 |
.03 |
27 |
n.s. |
Patient's ability to care for self |
|
|
|
|
|
Not able to take care of self |
75.9 |
3 |
-.13 |
31 |
n.s. |
No adequate place to live |
39.4 |
3 |
-.13 |
29 |
n.s. |
Not capable of working |
38.3 |
16 |
.47 |
31 |
p = .003 |
Family or friends favor commitment |
23.0 |
19 |
.39 |
23 |
p = .03 |
Predictions regarding patient |
|
|
|
|
|
Would not be a reliable outpatient |
82.0 |
2 |
.26 |
31 |
n.s. |
Cannot be counted on to take medications |
68.7 |
5 |
.18 |
31 |
n.s. |
Poor prognosis |
67.4 |
6 |
.15 |
31 |
n.s. |
Appropriate treatment not available at institution |
58.8 |
9 |
.01 |
31 |
n.s. |
Judge's information |
|
|
|
|
|
Expert witness (psychiatrist) not convincing |
85.3 |
1 |
-.19 |
31 |
n.s. |
Patient not well-known to judge |
-8.7 |
26 |
.01 |
31 |
n.s. |
Judge's opinion about patient's state |
|
|
|
|
|
Not in distress |
8.2 |
21 |
-.25 |
31 |
n.s. |
Composed |
26.9 |
17 |
-.43 |
31 |
p = .007 |
Cooperative |
2.6 |
22 |
.37 |
29 |
p = .02 |
Denies illness |
41.2 |
14 |
.39 |
29 |
p = .01 |
Seems incompetent |
24.1 |
18 |
— |
— |
— |
Unpredictable |
48.9 |
13 |
.53 |
31 |
p = .001 |
Judge's reaction to patient |
|
|
|
|
|
Appears well-groomed |
-4.0 |
25 |
-.18 |
31 |
n.s. |
Not frightening |
9.7 |
20 |
-.21 |
31 |
n.s. |
Not likable |
0.0 |
24 |
.47 |
26 |
p = .003 |
|
a-100 = strongly favoring release; +100
= strongly favoring commitment |
Discussion
The data revealed in this preliminary study raise important questions
for further investigation. The fact that almost all of the petitions
for civil commitment were granted—none was denied outright—is
striking. It should be seen, however, against a background of predominantly
voluntary hospitalization for mental illness. Commitment is a low-frequency
event. Unfortunately, the Department of Mental Health does not keep actuarial
data that would allow us to determine the actual ratio of committed (other
than on admission) to voluntarily hospitalized patients.
The high rate of commitments in our sample lends itself to several possible
explanations. Our sample of respondents may be skewed; that is, judges
who agree to participate in a study such as this may be those who are
most likely to accept the reasoning of a psychiatrist who petitions for
commitment. More generally, when a psychiatrist’s petition for commitment
of a patient who has filed a 3-day paper eventuates in the patient’s
release, it is frequently not by way of the judge’s denial of the petition.
Rather, during the typical week-long delay between the petition and the
hearing, the patient’s condition or the state of the doctor-patient relationship
usually improves to the point where the psychiatrist may withdraw the
petition or the patient may retract the request for discharge that precipitated
the commitment process. The cases actually heard in court, then, represent
a selected sample of intractable pathologies and irreconcilable differences.
Even so, our data do call into question whether the two-step commitment
process is functioning as it should. Given the low frequency of petitions
for commitment, psychiatrists appear to be setting too high a threshold
for seeking commitment. They appear to be petitioning only when they
feel sure that a judge will have solid grounds to commit rather than
when there are good clinical reasons for bringing up the matter for resolution
by the courts. Judges, probably correctly, sense this preselection of
committable patients, which thus dominates the decision-making process
and in practice forecloses the possibility of any outcome other than
commitment.
Such second-guessing is understandable: clinicians generally dislike going
to court and fighting for retention of patients and are thus likely to
petition only when success is almost certain. This approach, however,
prevents judges from becoming "calibrated" to degrees of dangerousness;
the entire process suffers from loss of the socially valuable heuristic
effect of judges grappling with hard cases.
Whether it is the judge who is swayed by psychiatric expertise or the
psychiatrist who allows legalism to inhibit clinical judgment, a pattern
of implicit collusion is suggested which resembles that found between
probation officers and judges in sentencing decisions [17].
Additional support for this line of reasoning may be drawn, for example,
from an unpublished 1984 study reported by Doucette et al. regarding
probable cause hearings in relation to commitment in California. The
authors noted,
"A significantly large proportion of patients were found to meet
the involuntary commitment criteria by the probable hearing officer,
suggesting that hospital treating staff were judicious in their certification
of individuals." We suggest that this "large proportion" reflects
the same anticipation our psychiatrists showed of the population likely
to be committed by the judges.
Further study drawing on larger and more representative samples of commitment
hearings is needed to distinguish between this and more benign explanations
of the high level of clinical-judicial agreement found here. In addition,
the judges’ reliance on the credibility of psychiatric testimony and
on predicted patient compliance with an outpatient regimen indicates
the value of ongoing feedback on outcomes following commitment or release
[1]. Finally, it would be informative in future studies
to have patient characteristics rated by disinterested observers and
then to correlate these ratings with the decisions of both psychiatrists
and judges.
It is important that experienced jurists, in their interpretation of legally
mandated criteria for commitment, resort to many of the same considerations
found relevant by experienced clinicians (our unpublished data). Indeed,
judges apparently keep in mind some of the same subjective factors when
making determinations of competence (in a committed population) that
clinicians consider when deciding whether a patient meets the legally
mandated criteria for commitment. Diagnosis plays a subsidiary role in
the judges’ assessments of competence. The positive correlation between
schizophrenia and competence appears attributable to the predominance
of paranoid schizophrenic patients seeking to leave the hospital. Of
the factors that correlated most strongly with competence, predictability
and capacity to work have clear clinical relevance. The unexpected correlation
between competence and likability may simply reflect the fact that judges
find it easier to identify with patients whom they perceive to be competent.
But judges, although denying by self-report the actual impact of such
factors on their decisions, may yet resemble clinicians in intuiting
that a patient who appears likable stands a better chance to have an
adequate support network outside the hospital than one who does not,
that competence may fluctuate according to interpersonal context, and
that loneliness may overwhelm the patient’s ability to execute wise decisions
[18].
We may question in retrospect how the judges understood the item referring
to competence when they judged virtually all of the patients in the study
to be committable, i.e., by inference incompetent to give informed consent
to hospitalization or discharge. Under these circumstances the judges
may simply have been noting degrees of competence within the patient
population to which they were exposed. Alternatively, they may have conscientiously
separated their commitment decisions from their estimates of a patient’s
capacity to make and implement life-sustaining choices. This implication
of the data should reassure those who are concerned that the determination
of committability and that of competence will be confounded under APA’s
model commitment statute [3], where incompetence is
made a criterion for committability.
Nonetheless, the judges in this study (like the authors of the model statute)
appear to have recognized a deeper confluence between the issues of committability
and competence. People who regularly come in contact with the mentally
ill, whether in a clinical or judicial capacity, realize that both direct
and indirect indicators of a person’s ability to function adequately
in a normal social environment must be taken into account in the commitment
decision. While following the letter of the law, judges necessarily interpret
it with the aid of relevant clinical cues reminiscent of those stressed
by the parens patriae model. Our study strongly suggests that
the argument between supporters of the parens patriae and "police
powers" models is far from resolved; both describe essential features
of a patient’s condition. Parens patriae may be legislated away,
but common sense and compassion alike keep the question of need for treatment,
occasionally clothed as competence, unmistakably in view.
References
-
Gutheil TG, Appelbaum PS: Clinical Handbook
of Psychiatry and the Law. New York, McGraw-Hill, 1982
-
Stone AA: Mental Health and Law: A System
in Transition. Rockville, Md, NIMH Center for Studies of Crime and
Delinquency, 1975
-
Stromberg CD, Stone AA: A model state law
on civil commitment of the mentally ill. Harvard Journal on Legislation
20:275-396, 1983
-
Appelbaum PS, Hamm RM: Decision to seek commitment:
psychiatric decision making in a legal context. Arch Gen Psychiatry
39:447-451, 1982
-
Cardozo BN: The Nature of the Judicial Process.
New Haven, Yale University Press, 1921
-
Rothman M, Dubin WR: Patients released after
psychiatric commitment evaluation: comparison with the committed.
J Clin Psychiatry 43:90-93, 1982
-
Ebbesen EB, Konecni VJ: Decision making and
information integration in the courts: the setting of bail. J Pers
Soc Psychol 32:805-821, 1975
-
Lelos D: Courtroom observation study of civil
commitment, in Civil Commitment and Social Policy. Edited by McGarry
AL, Schwitzgebel RK, Lipsitt PD, et aI. Rockville, Md, National Institute
of Mental Health, 1981
-
Hiday VA: Court discretion: application of
the dangerousness standard in civil commitment. Law and Human Behavior
5:275-289, 1981
-
Cohen F: The function of the attorney and
the commitment of the mentally ill. Texas Law Review 44:424-469,
1966
-
Rock RS, Jacobson MA, Janopaul RM: Hospitalization
and Discharge of the Mentally Ill. Chicago, University of Chicago
Press, 1968
-
Affieck GG, Peszke MA, Wintrob RM: Psychiatrists’
familiarity with legal statutes governing emergency involuntary hospitalization.
Am J Psychiatry 135:205-209, 1978
-
Massachusetts General Laws, chapter 123,
1, 11
-
Havens LL: Recognition of suicidal risks
through the psychologic examination. N Engl J Med 276:210-215, 1967
-
Motto JA, Heilbron DC, Juster RP, et al:
Suicide risk assessment: development of a clinical instrument, in
Proceedings of the 14th Annual Meeting of the American Association
of Suicidology. Denver, AAS, 1981
-
Fehrenback PA, O’Leary MR: Interpersonal
attraction and treatment decisions in inpatient and outpatient psychiatric
settings, in Basic Processes in Helping Relationships. Edited by
Wills TA. New York, Academic Press, 1982
-
Bankston WB: Legal and extralegal offender
traits and decision making in the criminal justice system. Sociological
Spectrum 3:1-18, 1983
-
Bursztajn H, Feinbloom RI, Hamm RM, et al: Medical
Choices, Medical Chances: How Patients, Families, and Physicians
Can Cope With Uncertainty. New York, Merloyd Lawrence, 1981