Dr. Appelbaum, who is editor of this column, is professor and director of the Division of Psychiatry, Law, and Ethics of the Department of Psychiatry, Columbia University College of Physicians and Surgeons. Address correspondence to Dr. Appelbaum at Department of Psychiatry, Columbia University Medical Center, 1051 Riverside Drive, #122, New York, New York 10032.

Jordan Nott, a student at George Washington University (GWU) in the nation's capital, must have thought he was falling through the looking glass. Feeling particularly depressed one October morning in 2004, and thinking about a friend who had committed suicide the previous spring, he asked two friends to take him to GWU Hospital, where he admitted himself to the psychiatric unit. Later that day, Nott received a letter from the assistant dean of students informing him that under the school's "psychological distress policy" he would not be permitted to return to his dormitory until he was cleared by the University Counseling Center and an administrative body. More was to come.

The following day, still hospitalized, Nott received a second, hand-delivered letter, this time from the dean of students herself. It charged him with a violation of the "endangering behavior policy" of the school's Residential Community Conduct Guidelines by virtue of his alleged suicidality. (Although admittedly depressed, Nott denies that he was suicidal or threatened to commit suicide.) According to the guidelines, "Behavior of any kind that imperils or jeopardizes the health or safety of any person or persons is prohibited. This includes any actions that are endangering to self or others." Nott was immediately suspended from the school, barred from all university property—including his dorm room—and told that he would be trespassing and could be arrested if found on campus after discharge from the hospital.

A hearing was scheduled the following week before GWU's judicial hearing board on his alleged violation of the guidelines. However, if he agreed to withdraw from the university, the "charges" against him would be deferred pending documentation that he had been symptom free for six months and was capable of living on his own. If Nott decided to go ahead with the hearing, he could be formally suspended or expelled. Ultimately, after reportedly being informed by a university official that he had little chance of succeeding at a hearing, Nott left GWU and enrolled at the University of Maryland, where he is currently a student. His father and brother emptied out his dorm room, while he sat in a car nearby, lest he be arrested for setting foot on university property. To this day, Nott is barred from the GWU campus (1,2,3).

What could have led a university to adopt a policy that makes suicidality a violation of its disciplinary code, an approach that the editorial page of the Washington Post aptly characterized as, "Depressed? Get Out!" (4)? It is worth noting that mandatory withdrawal for students who manifest suicidal ideation is neither unique to GWU nor particularly new (5,6,7). But there is some reason to believe that the incidence of such policies may be increasing and that the increase may be linked to administrators' fears of adverse publicity and legal liability if students commit suicide on campus (6).

Data on the incidence of suicidality on college campuses suggest one difficulty with such automatic responses. To be sure, college student suicide is a high-profile issue—suicide ranks second behind accidents as a cause of death among college students (8). Although good numbers are hard to come by, it is estimated that about 1,100 college students commit suicide in the United States each year (8); only a minority of students who attempt or complete suicide receive any treatment from college mental health services (9,10). The best epidemiologic data indicate that the rate of suicide on campus is about 7.5 of 100,000 students, roughly half of the expected rate in an age-matched community sample (11). However, with off-campus suicides difficult to tally, and students who are sent home from school or who leave voluntarily frequently missing from the statistics, the actual rate is hard to determine.

No matter how uncommon completed suicides are among college students, surveys suggest that suicidal ideation and attempts are remarkably prevalent. Two large-scale studies generated nearly identical findings (11,12). Roughly 10 percent of college student respondents indicated that they had thought about suicide in the past year, and 1.5 percent admitted to having made a suicide attempt. Combining data from the available studies suggests that the odds that a student with suicidal ideation will actually commit suicide are 1,000 to 1. Thus policies that impose restrictions on students who manifest suicidal ideation will sweep in 999 students who would not commit suicide for every student who will end his or her life—with no guarantee that the intervention will actually reduce the risk of suicide in this vulnerable group. And even if such restrictions were limited to students who actually attempt suicide, the odds are around 200 to 1 against the school's having acted to prevent a suicidal outcome.

Two recent legal cases, however, have created a climate of fear among administrators. In Schieszler v. Ferrum College, a federal district court applying Virginia law found that a college and its staff members who knew that a student was manifesting suicidal ideation such that there was "an imminent probability" of harm could be held liable for failing to act to prevent his death (13). A similar ruling came from a Massachusetts court in Shin v. Massachusetts Institute of Technology, a case that was the subject of an extended profile in the New York Times Magazine (14). Although the court dismissed the Massachusetts Institute of Technology as a defendant in the case, holding that it had not had a responsibility to prevent a student, Elizabeth Shin, from immolating herself in her dormitory room, the ruling allowed the case to proceed against both the medical professionals who had treated her at the university health service and two administrators who had been involved with her. As in Schieszler, the court found that the administrators had a "special relationship" with Shin that created a legal duty to protect her from harm, based on their presumed ability to foresee that she would hurt herself without proper supervision (15).

Together, Schieszler and Shin (which was recently settled out of court for an undisclosed amount [16]) have left many college administrators fearful of their liability should any suicide occur on campus. This response, though understandable, may not be entirely reasonable, given that most courts have declined to impose liability on an educational institution or its agents in the wake of a student's self-inflicted death (17). Administrative responses have ranged from refusing to allow the gathering of identifiable information concerning students who manifest suicidality—for fear that this knowledge would provoke a corresponding duty to protect them—to beefed-up policies requiring mandatory leaves of absence and forcing students with suicidal ideation to sign agreements that they will talk with no one on campus about their problems (6,18).

However, liability in this context does not come from a single direction. Jordan Nott has sued GWU, alleging deprivation of his rights under the Americans With Disabilities Act, section 504 of the Rehabilitation Act, the Fair Housing Act, and several District of Columbia statutes, as well as intentional infliction of emotional distress. The Bazelon Center for Mental Health Law, which seeks to protect the rights of persons with mental disorders, has provided the lead counsel in the case. In addition, Nott has sued his caregivers and GWU Hospital for breaching the privacy of his medical information by allegedly disclosing information about his hospitalization and condition to the university (1). The university has defended its actions but has not publicly challenged Nott's account of the case, which has not yet come to trial.

In addition to Nott's suit, a number of complaints have been filed with the federal Office of Civil Rights (OCR) over mandatory leave policies, and in four recent cases colleges were found to have violated students' rights under section 504 of the Rehabilitation Act and were ordered to readmit them (6). In one case, involving Bluffton University in Ohio, OCR ruled that the institution had acted on the basis of the mental illness manifested by a student who had attempted suicide and had failed to provide a hearing, consult with medical personnel, examine objective evidence, or evaluate the real nature and extent of the risk she posed to herself or others. The school was forced to change its policy and reimbursed the student for her room, board, and book fees.

In this complicated legal context, what would constitute reasonable policy for a college or university faced with a student who has manifested suicidal ideation or behavior? There is no question that suicidal students can be disruptive to college life, as when dormitory mates feel compelled to stay up with a distraught friend all night to keep the person safe. Moreover, contagion of suicide appears to be a real phenomenon, with some campuses experiencing mini "epidemics" of self-harm. So schools have some legitimate interest in ensuring that students receive appropriate treatment, and there are undoubtedly cases in which that can best be accomplished by means of an extended leave.

However, there are better and worse ways for colleges to protect their interests. Categorizing suicidality as an infraction of the school's disciplinary rules creates the Kafkaesque situation in which depressed students must defend themselves at a hearing against the "charge" of having tried to hurt themselves—and seems likely to dissuade depressed students from seeking treatment. Responses to suicidal ideation and behavior are better addressed as part of a school's medical policies. Given the diversity of situations in which students may manifest suicidality, a blanket policy requiring withdrawal from campus for a specified period is also not appropriate. Schools may benefit from not having to treat depressed students in their health services and from avoiding the opprobrium of an on-campus suicide in the event of a bad outcome. However, being placed on leave may not be in the best interests of depressed students, who may thereby be separated from their social network and lose the self-esteem that comes from success in academic and extracurricular activities. The assumption that sending a student home will reduce the stress they experience is naïve at best, given that home may be precisely the source of conflict in the student's life.

Thus decisions about whether to place a student on leave should be made on an individual basis, with the presumption that a student will continue in school if he or she chooses. So as not to run afoul of students' rights under federal disability law, decisions should be made on the basis of defined procedures and criteria, and the student should have the opportunity to present evidence, including psychiatric evaluations. Treatment by the university counseling service or its equivalent should remain confidential and entirely separate from whatever administrative process exists for decisions about leave and readmission—which is not always the case today. Colleges and universities have to find other ways of minimizing their perceived legal and public relations risks than acting in a manner that adversely affects the rights and interests of students who may suffer from depression and suicidality.