MENTAL INSTITUTIONS, COMMITMENT TO

Throughout the world there are legal mechanisms by which mentally ill persons can be sent to psychiatric hospitals even when they do not wish to go (Appelbaum). In the United States this sometimes is done through the criminal justice system: A person may be judged incompetent to stand trial for a crime because of mental illness or may be tried for a crime and found not guilty by reason of insanity and then committed to an institution for mentally ill criminal offenders. The more common type of commitment is civil, and usually no criminal offense is involved: A person is judged to require hospitalization because of his or her mental condition but does not consent to it, but if certain legal criteria are fulfilled, that person may be hospitalized against his or her will. Commitment is a legal process and often is discussed mainly in terms of its case and statutory legal history (Wexler). This entry discusses important ethical issues that underlie the process of civil commitment.

Commitment raises serious ethical concerns. It involves depriving persons of their freedom for days, weeks, or longer, usually by incarcerating them in a locked psychiatric facility. Commitment is one of the ethically most serious actions in which psychiatrists engage. However, neither the process of commitment nor its ethical justification (or the related issue of forced treatment) is mentioned in the American Psychiatric Association's extensive handbook on psychiatric ethics (American Psychiatric Association, 2001b).

In most states this violation of a person's civil liberties can be carried out initially on an emergency basis on the strength of one physician's signature on the appropriate form. Most people agree that it is preferable that a psychiatrist be the initial committing physician, but there are too few psychiatrists in many rural areas for this usually to be mandated by law.

After the emergency commitment form is signed, the person who is to be committed is taken to the nearest locked psychiatric facility authorized to receive committed persons. Medical personnel there usually have the authority to question the appropriateness of the commitment and even to refuse to detain the person. In most states, under modern law, a probable-cause judicial hearing is held within two to three working days in an appropriate local court to determine the justifiability of continued detainment.

The vast majority of admissions to psychiatric hospitals, however, are voluntary and do not involve the commitment process. A small minority of voluntary admissions, however, result from persons being told that they will be committed if they do not enter the hospital "voluntarily." There seems to be nothing inherently unethical about giving a person who otherwise would be committed the opportunity to avoid the commitment process in that way, assuming that the planned commitment is ethically justified. It seems clear, however, that these persons have not entered the hospital entirely voluntarily. In addition, it would be prima facie unethical for a physician to use this process deceptively by manipulating a person into entering a hospital by threatening a commitment that in fact would not be carried out.

Legal Criteria for Commitment

Both within and outside psychiatry there is dispute about the commitment criteria that should be written into state statutes. Statutory language varies from state to state (Arthur et al.). All U.S. state statutes stipulate that to be committed a person must be mentally ill, although this concept is defined variously. The existing continuum of positions is based on the width or narrowness of the additional statutory commitment criteria. (For an excellent discussion of one state's commitment laws see Behnke, Winick, and Perez.)

The broadest additional criteria are advocated by those who think that physicians should be able to commit anyone whom they sincerely believe would profit from commitment. At one time many states had statutes with this breadth. Arizona law, for example, as recently as 1981 allowed persons to be detained if they were "mentally ill and in need of supervision, care or treatment" (Wexler, p. 74). Criteria with this breadth seem unsupportable to most commentators. For example, many persons with a moderate degree of depression are mentally ill in that they satisfy the criteria in the Diagnostic and Statistical Manual, Fourth Edition (DSM-IV) (American Psychiatric Association, 1994) for having a psychiatric disorder, and treatment almost certainly would make them feel better. No one, however, thinks that in most cases they should be forced into a psychiatric hospital if they do not wish to go. Thus, more than mental illness is necessary to justify commitment.

A narrower position is taken by many psychiatrists (see Chodoff for a classic description of this position and Buchanan and Brock for clear arguments supporting it). In addition to requiring that a person be mentally ill, supporters of this position advocate a criterion stipulating that that person be gravely disabled or manifest a serious disruption of functioning as a result of the mental illness. Being physically dangerous to oneself (suicidal) or to others (homicidal or physically threatening) represents one type of serious disruption of functioning but not the only one. The behavioral and social disorganization shown by many manic persons, for example, although often not immediately physically threatening to themselves or to others, may in the long run cause those persons serious social and financial harm. Under a serious disruption criterion many of those individuals could be committed.

A narrower position still is that advocated by many civil libertarians and some psychiatrists (American Bar Association). A diagnosis of mental illness is required, and there must be a high probability that because of the mental illness a person is a serious physical threat to himself or herself or to others. A minority in this group would restrict the criterion further and require that there be good evidence of recent behavior toward oneself or others that was in fact physically harmful, but most believe that evidence of strong threats of physical harm is sufficient. Most also believe that dangerousness toward oneself can be evidenced not only by threats of suicide but also by extreme self-neglect so that, for example, starvation or untreated serious disease can constitute an immediate threat. However, without the threat of imminent dangerousness of some kind, commitment would not be allowed.

The position at the far end of the continuum is taken by those who believe that psychiatric commitment is never ethically justified and thus that there should be no commitment criteria. Thomas Szasz, a psychiatrist, has been the foremost spokesperson for this position. Szasz believes that the concept of mental illness is mythical and argues that those who manifest what others regard as the symptoms of mental illness should be judged only by the standards of criminal law: If they have broken a law, they may be arrested or otherwise constrained; if they have not, their freedom should be preserved. Szasz believes that commitment is based on a false theory that "medicalizes" deviant behavior into illness and that psychiatrists who commit persons become unwitting arms of the criminal justice system.

For several reasons Szasz's position has not been persuasive to many people inside or outside psychiatry, including most civil libertarians. First, most scholars feel that some psychological conditions satisfy the criteria of a definition of illness (Gert, Culver, and Clouser, Margolis) and that Szasz's position has serious theoretical problems (Moore, Culver, and Gert) that he has not addressed. Second and more important, most believe that paternalistic interventions of the type that commitment usually represents are at least sometimes ethically justified.

The principal and enduring tension is between those who hold the two middle positions described above. Some states have commitment statutes closer to one, and some have statutes closer to the other. Those who advocate a broader criterion believe that dangerousness to oneself and others is only one of many manifestations of severe mental illness and that it is cruel and theoretically unjustifiable to ignore the needs of disordered or disabled persons, often homeless and wandering the streets, who clearly would benefit from treatment (Treffert, Peele, and Chodoff; American Psychiatric Association, 2001a). References are made to people "dying with their rights on" and to Janis Joplin's song line "Freedom's just another word for nothin' left to lose."

Those who advocate the narrower grounds fear that relaxing the criterion in the direction of disruption of functioning leaves the door open too wide to psychiatric paternalism and represents a threat to civil liberties. Images of forced psychiatric internment of political dissidents in the Soviet Union (Bloch and Reddaway) are invoked as a frightening example of giving psychiatrists the power to confine individuals who are not physically dangerous but only disrupted in their functioning. One of the necessary and willing prices of having a free society, they argue, is that people are free to make self-defeating choices and sometimes irrationally reject opportunities for help.

A cohort of persons are committable under a broader but not under a narrower set of criteria. An example is a person with a history of bipolar disorder who becomes increasingly hypomanic and is squandering his carefully accumulated savings in what are almost certainly hopeless financial schemes. He refuses all treatment. Everyone who knows him believes that his spending spree is due to his hypomania, that it would not be unethical to curtail his actions, and that if his behavior were curtailed, he almost certainly would be grateful later. However, although his current behavior is harmful to his long-term interests, he is not dangerous to himself or others as that criterion is explicated in many states.

Many persons, like this man, whose behavior meets broader but not narrower commitment criteria suffer from cyclical disorders: Their aberrant behavior occurs only episodically. Some authors have suggested that such persons might be offered during nonsymptomatic times the opportunity to create a contract stating that if their future behavior deviates from their usual behavior in certain specified ways, they will accept the use of appropriate interventions (confiscation of funds or forced hospitalization, voluntary commitment) that otherwise might not be legally permissible (Howell et al., Culver and Gert).

An important empirical issue discussed by Peele and Chodoff is the extent to which statutory criteria for commitment influence the behavior of psychiatrists. Are there patients who are not committed in states with narrow criteria who would be committed in states with broader criteria? Peele and Chodoff, after surveying the scanty evidence that exists on this point, conclude, "It appears that judges and juries base decisions about commitment on what they think is best for the person, regardless of formal criteria" (Peele and Chodoff, p. 436). This would be a useful issue to explore further.

Conceptual Issues Underlying Commitment

ETHICAL JUSTIFICATION. In discussing the ethical justification of commitment a distinction must be made between whether a commitment is intended primarily to help the person who is committed or to help others whom that person may be putting at risk (Gert, Culver, and Clouser; Buchanan and Brock). This distinction sometimes is not clear-cut because it is usually to the advantage of mentally ill persons to be prevented from harming others. The harm they might cause often would be serious and thus would constitute a crime. Committing the crime frequently would be a clear result of the mental illness—for example, obeying a voice commanding that someone be killed—and it is highly likely that the mentally ill offender would be apprehended, incarcerated, and then punished or at least hospitalized for a long time. Nonetheless, there is a distinction between paternalistic and nonpaternalistic commitments, and there is no doubt that the protection of others is the predominant reason for some commitments.

Paternalistic commitment. To the extent that commitment is intended to help the person who is committed, it essentially always qualifies as a paternalistic action. That is, the commitment is intended to benefit the committed person, it violates at least one moral rule (deprivation of freedom) and usually several, it is done without the consent of the person, and the person is at least minimally competent to give consent (Gert, Culver, and Clouser). Whether paternalistic commitment is ethically justified therefore depends on whether a particular commitment meets whatever theoretical criteria for justified paternalism are thought to be adequate.

Various sets of criteria, partly overlapping, have been proposed by Beauchamp and Childress, Buchanan and Brock, Childress, and Gert, Culver, and Clouser. Those criteria depend on theoretical concepts such as the degree of irrationality and voluntariness of the person's behavior and the balancing of physician beneficence and patient autonomy. None of those authors seems to believe that as a species of paternalism, there is anything qualitatively unique about committing mentally ill individuals. Thus, particular acts of commitment are measured directly against the theoretical criteria of the particular justification procedure that is proposed.

However, in the judgment of many authors (Culver and Gert; Buchanan and Brock), the presence of mental illness does play an indirect role in the justification of paternalistic commitment by sometimes affecting concepts that those authors believe are centrally important in the justification process. Thus, some suicidal desires may be regarded as not truly expressing an individual's autonomous wishes (Beauchamp and Childress), or some conditions of mental illness may be thought to affect a person's competence to make decisions (Buchanan and Brock).

Nonpaternalistic commitment. When commitment is not paternalistic, it must be ethically justified on other grounds. To commit persons in an attempt to prevent them from harming others represents a kind of preventive detention that ordinarily is not legally permitted in the United States. In the presence of some kinds of mental illness, however, it is argued by some that nonpaternalistic commitment may be ethically justified.

For example, two men are brought separately to the emergency room by the police. In each instance the police have been called because the man has just threatened to kill his wife. Each man admits to the emergency room psychiatrist that this is true. The first man has a history of paranoid psychotic episodes and in recent days has heard voices instructing him to kill his wife. The second man has no symptoms or history of major mental illness, but he and his wife have a history of chronic marital discord. In both cases the psychiatrist feels that there is a reasonably high probability that the man will harm his wife if he returns home.

On the basis of the fact that in some kinds of mental illnesses persons are not held responsible for their actions, it may be argued that it is ethically justified to commit the first man but not the second. The second man, for example, presumably has the volitional ability to will or to refrain from willing to harm his wife, whereas the first may not have the volitional ability to will not to harm her (Culver and Gert). Dangerous mentally ill persons sometimes are not considered capable of guiding their behavior in accordance with promulgated social rules (Brock).

PREDICTING POSSIBLE FUTURE HARM. Civil commitment always involves a doctor's appraising a person's physical and mental status and deciding whether commitment is warranted. Sometimes individuals may be committed because they are in such a disabled condition that even more serious future harm seems all but inevitable. A woman may, for example, be hallucinating continuously, be unresponsive to the questions or actions of others, and be significantly malnourished because of a lack of interest in food. Much more often, however, serious future harm is only a possibility: For example, a person has threatened suicide or is hearing voices urging her to harm someone, and the physician must try to predict how likely it is that the harm actually will occur.

The process of predicting possible future harm in the commitment setting has the following components (Grisso): The criterion is what is being predicted (for example, the person's suicide), the cues are discrete pieces of available information about a particular case at a particular point in time (for example, the person's age, sex, state of intoxication, and history of impulsivity), and the judgment is the physician's conclusion after assessing the case (for example, to commit or not to commit). These are three separate elements. Empirical research has focused separately on the correlations among them. The judgment-criterion correlation shows how well physicians do in predicting that particular persons will kill themselves. The cues-criterion correlation shows the extent to which suicides can be predicted from whatever facts about cases can be isolated and measured independently of physicians' judgments. The cuesjudgment correlation shows which data about cases lead physicians to make one judgment or another.

A critically important issue with respect to prediction is the extent to which commitment does prevent future serious harm. There are few data addressing this issue. If it were known, for example, that 90 percent of the persons committed would have harmed themselves or others seriously if they had not been committed, most people probably would feel that commitment was ethically justified. Committing one hundred persons would avoid ninety instances of serious harm, although at the cost of committing ten persons who would not have caused harm if they had not been committed. By contrast, if only one in a hundred persons would have harmed themselves or others, few would feel commitment was justified because ninety-nine persons would have suffered the evils of detainment to prevent one bad future outcome.

This kind of utilitarian calculus seems central to most writers who discuss the ethical justifiability of commitment. Commitment essentially always inflicts significant harm, but only sometimes does it prevent significant harm. Almost everyone acknowledges that even among those at relatively high risk of causing harm—for example, suicidal persons brought to an emergency room—only a minority would, if left alone, subsequently harm themselves. An emergency room physician thus faces a difficult task. To commit every person would be to commit too many, but which persons should be committed? Certain characteristics of persons (cues) are known to increase the likelihood of future harmful acts—for example, a history of impulsive or suicidal behavior, being inebriated, having access to lethal weapons, being male—but a physician must make a binary, yes-no decision about commitment, not a probability estimate.

Research (Monahan) suggests that physicians are poor predictors of whether harmful behavior will occur (judgmentcriterion correlations). There is reason to believe that basing predictions on discrete, measurable pieces of information about a case (cues-criterion correlations) will yield greater accuracy (Monahan). There is, however, probably an upper limit to predictive accuracy; one reason for this is that whether a person commits a harmful act in the hours or days after a physician's assessment may depend at least as much on later fortuitous situational factors such as whether a friend returns a telephone call as on factors that can be measured during the assessment.

A very important statistical feature of prediction plays a key role in understanding the commitment process and making ethical judgements about it. In predicting relatively rare events such as the occurrence of a future suicide through the use of predictive signs of less than extremely high predictive accuracy (for example, a physician's judgment or whether a person has access to a lethal weapon), one inevitably will make a high proportion of false-positive predictions; that is, one frequently will predict future harm when in fact none will occur. This actuarial problem, which is an example of the application of Bayes' theorem, was described by Meehl and Rosen and later applied to the issue of commitment by Livermore, Malmquist, and Meehl.

Suppose that 10 percent of suicidal persons who are brought to an emergency room but are unwilling to be hospitalized would kill or harm themselves seriously if they were not committed. Suppose further that, using the available cues, physicians' predictions of who will and will not commit suicide have a sensitivity of 70 percent (sensitivity refers to the percentage of persons who will commit suicide whom physicians accurately predict will commit suicide) and a specificity of 70 percent (specificity refers to the percentage of patients who will not commit suicide whom physicians accurately predict will not commit suicide). It follows that physicians will commit and thus save seven of the ten persons destined for suicide but also will commit twenty-seven persons of every ninety persons (30% of ninety) who would not have killed themselves. These latter persons constitute false positives.

The ratio between the number of true positives (seven) and false positives (twenty-seven) shows that nearly four persons will be committed needlessly in order to save one. (These are hypothetical figures. Many would argue that subsequent suicide is rarer than 10 percent in the general psychiatric suicidal population and that 70 percent is too high an estimate of sensitivity (and of specificity); thus, the actual proportion of false positives would be much higher.) The physician would be correct a higher percentage of the time (90%) if he or she simply predicted that no one would commit suicide, but then none of the ten suicidal persons would be saved.

Is it ethically justified to commit four unwilling persons needlessly to save one life? Suppose empirical data existed (they do not) that enabled the construction of actuarial tables that would correlate the nature and number of signs and symptoms shown by mentally ill persons in emergency rooms with their subsequent likelihood of harming themselves or others if they were not committed (cue-criterion correlations). Each person thus could be assigned to a cohort: Some would have a one in five chance of harming themselves or others, some a one in ten chance, some one in twenty, some one in forty, and so forth.

Where should the line be drawn? What is the appropriate trade-off between saving one life and needlessly depriving many persons of their freedom? Reasonable people might disagree about where the line should be drawn, but this is a matter that could be opened to public debate. Psychiatrists probably have no special expertise in deciding where the threshold for commitment should be placed.

When confronted with the inevitable large numbers of false-positive commitments, some people recall the injunction often cited in connection with the U.S. criminal justice system—"Better that ten guilty persons go free than one innocent person suffer"—and conclude that civil commitment is ethically unjustified (Sartorius). Others, however, although concerned about the false-positive problem, believe that there are sufficient differences between the underlying conceptual justifications of the criminal justice system and the civil commitment system that some number of false positives can be tolerated in the civil system (Brock).

Conclusion

Although debates about involuntary hospitalization sometimes are framed in legal rather than ethical terms, it is important to be clear about the underlying ethical issues. Civil commitment involves incarcerating an unwilling person who has committed no crime for days, weeks, or longer. This type of prima facie unethical action requires clear justification in terms of a general moral theory. Current theoretical discussions of commitment emphasize concepts such as the degree of irrationality and the extent of voluntariness of a person's behavior. In applying theoretical concepts to the process of commitment it is critical to describe the components of the process clearly and take into account certain statistical features that are inherent in making predictions about a person's future behavior.

charles m. culver (1995)

revised by author

SEE ALSO: Autonomy; Behavior Control; Coercion; Competence; Human Rights; Institutionalization and Deinstitutionalization; Mental Illness; Mentally Disabled and Mentally Ill Persons; Patients' Rights: Mental Patients' Rights

BIBLIOGRAPHY

American Bar Association. 1977. "Legal Issues in State Mental Health Care: Proposals for Change." Mental Disability Law Reporter 2(1): 57–159.

American Psychiatric Association. 1994. Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition Washington, D.C.: Author.

American Psychiatric Association. 2001a. Ethics Primer. Washington, D.C.: Author.

American Psychiatric Association. 2001b. Principles of Medical Ethics with Annotations Especially Applicable to Psychiatry. Washington, D.C.: Author.

Appelbaum, Paul. 1997. "Almost a Revolution: An International Perspective on Involuntary Commitment." Journal of the Academy of Psychiatry and Law 25: 135–148.

Arthur, Lindsay G.; Haimovitz, Stephan; Lockwood, Robert W.; Dooley, Jeanne A.; and Parry, John W., eds. 1988. Involuntary Civil Commitment: A Manual for Lawyers and Judges. Washington, D.C.: American Bar Association.

Beauchamp, Tom L., and Childress, James F. 2001. Principles of Biomedical Ethics, 5th edition. New York: Oxford University Press.

Behnke, Stephen H.; Winick, Bruce J.; and Perez, Alina M. 2000. The Essentials of Florida Mental Health Law. New York: Norton.

Bloch, Sidney, and Reddaway, Paul. 1977. Psychiatric Terror. New York: Basic Books.

Brock, Dan W. 1980. "Involuntary Civil Commitment: The Moral Issues." In Mental Illness: Law and Public Policy, ed. Baruch A. Brody and H. Tristram Engelhardt, Jr. Boston: D. Reidel.

Buchanan, Allen E., and Brock, Dan W. 1989. Deciding for Others: The Ethics of Surrogate Decision Making. New York: Cambridge University Press.

Childress, James F. 1982. Who Should Decide? Paternalism in Health Care. New York: Oxford University Press.

Chodoff, Paul. 1976. "The Case for Involuntary Hospitalization of the Mentally Ill." American Journal of Psychiatry 133(5): 496–501.

Culver, Charles M. 1991. "Health Care Ethics and Mental Health Law." In Law and Mental Health: Major Developments and Research Needs, ed. Saleem A. Shah and Bruce D. Sales. Rockville, MD: U.S. Department of Health and Human Services.

Culver, Charles M., and Gert, Bernard. 1982. Philosophy in Medicine: Conceptual and Ethical Issues in Medicine and Psychiatry. New York: Oxford University Press.

Gert, Bernard; Culver, Charles M.; and Clouser, K. Danner. 1997. Bioethics: A Return to Fundamentals. New York: Oxford University Press.

Grisso, Thomas. 1991. "Clinical Assessments for Legal Decision Making: Research Recommendations." In Law and Mental Health: Major Developments and Research Needs, ed. Saleem A. Shah and Bruce D. Sales. Rockville, MD: U.S. Department of Health and Human Services.

Howell, Timothy; Diamond, Ronald J.; and Wikler, Daniel. 1982. "Is There a Case for Voluntary Commitment?" In Contemporary Issues in Bioethics, 2nd edition, ed. Tom L. Beauchamp and LeRoy Walters. Belmont, CA: Wadsworth.

Livermore, Joseph M.; Malmquist, Carl P.; and Meehl, Paul E. 1968. "On the Justifications for Civil Commitment." University of Pennsylvania Law Review 117(1): 75–96.

Margolis, Joseph. 1976. "The Concept of Disease." Journal of Medicine and Philosophy 1(3): 238–255.

Meehl, Paul E., and Rosen, Albert. 1955. "Antecedent Probability and the Efficiency of Psychometric Signs, Patterns, and Cutting Scores." Psychological Bulletin 52(3): 194–216.

Monahan, John. 1984. "The Prediction of Violent Behavior: Toward a Second Generation of Theory and Policy." American Journal of Psychiatry 141(1): 10–15.

Moore, Michael S. 1975. "Some Myths about 'Mental Illness.'" Archives of General Psychiatry 32(12): 1483–1497.

Peele, Roger, and Chodoff, Paul. 1999. "The Ethics of Involuntary Treatment and Deinstitutionalization." In Psychiatric Ethics, ed. Sidney Bloch, Paul Chodoff, and Stephen A. Green. New York: Oxford University Press.

Sartorius, Rolf E. 1980. "Paternalistic Grounds for Involuntary Civil Commitment: A Utilitarian Perspective." In Mental Illness: Law and Public Policy, ed. Baruch A. Brody and H. Tristram Engelhardt, Jr. Boston: D. Reidel.

Szasz, Thomas S. 1970. Ideology and Insanity: Essays on the Psychiatric Dehumanization of Man. Garden City, NY: Anchor.

Treffert, Darold A. 1985. "The Obviously Ill Patient in Need of Treatment: A Fourth Standard for Civil Commitment." Hospital and Community Psychiatry 36(3): 259–264.

Wexler, David B. 1981. Mental Health Law: Major Issues. New York: Plenum.