November 28, 2022


Arnold R. Isaacs


Communication. Does that single word define the path to safer college and university campuses? Deliberations at a two-day meeting organized by the Academy for Critical Incident Analysis suggest that among those responsible for campus safety and students’ well-being, many would answer yes — that better communication is, indeed, the key to more effective violence prevention and, consequently, a lower risk of harm to students, faculty and staff at American institutions of higher education. The meeting, held March 11-12 on the Columbia University campus, brought together administrators, directors of student life and counseling services, and public safety officials from a half-dozen institutions in the New York area, along with a number of other experts and observers. Virginia Tech University was also represented, as the site of the deadliest campus shooting in U.S. history and also as the subject of ACIA’s last case-study conference, held in July, 2009, in Blacksburg, Virginia.


Balancing Confidentiality and Safety


The Columbia conference was explicitly organized to follow up on a key issue raised by the Virginia Tech shootings: whether the shooter’s mental health history and disturbing behavior should have raised warning flags that might, if handled differently, have made it possible to prevent the tragedy. The same question has arisen after other violent events, in particular the 1999 shooting at Columbine High School in Littleton, Colorado (which, incidentally, appears to have had a significant influence on Seung Hui Cho, the Virginia Tech shooter).

Information-sharing issues were pinpointed by the Virginia Tech Review Panel, which noted in its summary of key findings that during Cho’s junior year on campus, “numerous incidents occurred that were clear warnings of mental instability. Although various individuals and departments within the university knew about each of these incidents, the university did not intervene effectively. No one knew all the information and no one connected all the dots.” In the end, the panel continued, “the system failed for lack of resources, incorrect interpretation of privacy laws, and passivity.” In the body of the report, the panel amplified on that finding: “the lack of information sharing among academic, administrative, and public safety entities at Virginia Tech and the students who had raised concerns about Cho contributed to the failure to see the big picture…. The Care Team was hampered by overly strict interpretations of federal and state privacy laws (acknowledged as being overly complex), a decentralized corporate university structure, and the absence of someone on the team who was experienced in threat assessment and knew to investigate the situation more broadly, checking for collateral information that would help determine if this individual truly posed a risk or not.”1


About the author


Arnold R. “Skip” Isaacs is an author, freelance writer and editor, and educator. He was previously a reporter, foreign and Washington correspondent, and editor for the Baltimore Sun. He is the author of two books relating to the Vietnam War, Without Honor: Defeat in Vietnam and Cambodia and Vietnam Shadows: The War, Its Ghosts, and Its Legacy.

Most campus public safety officials at the Columbia conference appeared to feel that in their institutions — possibly in part as a consequence of the Virginia Tech tragedy — barriers to communication are being overcome and information is getting to those who need it. “People realize lack of communication is going to kill us,” said Kenneth E. Finnegan, public safety director for Columbia’s Morningside campus, adding that information sharing has “improved dramatically, both from the mental health and the dean of student life sides of the house.” John DeAngelis, director of public safety at Columbia Teachers College, agreed: “I think we are in the best position we’ve ever been in… we’re in better shape than in years past.” A factor in this improvement may have been new guidelines for colleges and universities issued by the U.S. Department of Education several months after the Virginia Tech shootings in an attempt to clarify what disclosures are and are not permitted by the Family Educational Rights and Privacy Act, commonly referred to as FERPA. The guidelines specifically authorized designating campus police as “school officials” with a “legitimate educational interest,” and thus able to receive “personally identifiable information from students’ education records.”2

There is, of course, a strong reason for protecting confidentiality of student information, mental health records in particular. Christopher Rodgers, dean of students at Fordham University’s Rose Hill campus, reminded the conference that the wall between a university administration and psychiatrists, psychologists and social workers exists and must be protected “so you can get students to come when they need to.” Without confidentiality safeguards, he added, “we are never going to hear half of things we hear about now.” Law, professional ethics, institutional policies and common sense all recognize that there are situations requiring exceptions to confidentiality rules, but there is no easy or simple formula determining what and how to disclose in a given case. That, as Dr. Paul S. Appelbaum pointed out in keynoting the March 12 session, is why “clinicians and administrators struggle constantly” with the issue — a necessary and appropriate struggle, because those decisions involve balancing different needs and obligations.

Appelbaum, professor of psychiatry, medicine and law in the psychiatry department of Columbia’s College of Physicians and Surgeons and director of the department’s division of law, ethics and psychiatry, commented that “we can learn from Virginia Tech and similar tragedies of lesser magnitude” where more disclosure might have helped avoid a disastrous event. But, he added, “rare events should not drive policy.” It is also worth remembering that universities and colleges are not dangerous places. Nor are they becoming more dangerous. The incidence of violent crime is much lower on campuses than in the nation as a whole, and available data show that rate declined, rather than rising, between the mid-’90s and the middle of the last decade. Though mass-casualty incidents like the shootings at Virginia Tech may heighten public fears, they do not represent an epidemic of campus violence or a need for drastic new limitations on confidentiality protection.

In view of the potential consequences for an individual and for the effectiveness of a counseling service, Appelbaum concluded, disclosure of mental health information should take place only when other options are unlikely to work, and specific information released should be tailored to needs of the situation. In most cases, he said, appropriate situations for disclosure will be those already listed in privacy laws as emergency exceptions — when third parties are in danger, in commitment proceedings, or when a patient is incompetent to consent to disclosure, among others. “Probably a reasonable approach for the institution as a whole,” he added, is the common-sense standard that disclosure “should do more good than harm.”


The Threat Assessment Model


Typically, prevention issues capture attention following a spectacular, high-casualty event such as the fatal rampages by deeply disturbed students at Columbine High or Virginia Tech. Obviously, the occurrence of such events means that prevention has failed. Examining those failures can lead to useful changes (the revised FERPA guidelines issued after the Virginia Tech shootings are one example). But concentrating too narrowly on what went wrong in a particular high-profile incident can also be a trap, for a number of reasons.

Mass shootings on campus are extremely rare — which means that looking too hard for similar warning flags may not be of significant help in preventing another incident because similar signals will in fact not reflect the same danger: the “false positive” problem. The writings and videos that reflected Seung Hui Cho’s and the Columbine killers’ violent fantasies were obvious warning signs, in retrospect. But the same kind of fantasies will be expressed by an overwhelmingly larger number of adolescents who will not become mass murderers. That does not mean that troubling works of imagination should be ignored. But it does mean that a student’s story or essay or film with disturbing visions of violence is not very useful, by itself, as a forecaster of a violent act. Nor are other apparent alarm signals taken by themselves, such as suicidal thoughts. Survey data consistently indicate that among students who report having seriously considered suicide, only one in every thousand will actually take his or her own life.

From the standpoint of prevention, the most important thing to do about such specific indicators (in the absence of an emergency situation requiring immediate response) is not to leap to intervene against the apparent — though frequently minimal — risk. More important is to find out whether they are associated with other indicators and circumstances. Has there been a broken romance or a family conflict? A history of threatening or violent acts in the past? Drug or alcohol abuse or other self-destructive behavior? A diagnosis of depression or other mental health disorder? A noticeable drop in grades or work performance?  Financial pressure or the threat of job loss or some other acute stress in the person’s life? Does the person have a weapon, or access to one? Is he or she familiar with firearms?

Such questions are at the core of the process called threat assessment, which has, in many institutions, become the centerpiece of violence prevention efforts. Gene Deisinger, deputy chief of the Virginia Tech Police and keynote speaker at the opening session of the Columbia conference, described threat assessment as a systematic, evidence-based process that moves through four stages: first, identifying “persons of concern”; second,  ”collecting information from multiple sources and constructing a much fuller picture of a person’s emotional condition and personal circumstances”; third, assessing that information and the situation; and fourth, managing the situation. Once the first two steps have been taken, Deisinger said, the next two — assessing and managing the threat — involve another four-part concept: that violence “is the product of an interaction among four factors” that he listed under the acronym STEP. S represents the Subject who may take violent action; T refers to the vulnerabilities of the Target of such actions; E stands for the Environment that may facilitate or permit or fail to discourage violence; and P indicates Precipitating events that may trigger an act by the subject.

Threat management, accordingly, deals with all four factors. It seeks, in Deisinger’s formulation, to “de-escalate, contain, or control the Subject who may take violent action”; to “decrease vulnerabilities of the Target”; to modify the physical and cultural Environment in ways that will discourage violence; and to prepare for and find ways to mitigate possible Precipitating events that may be the spark for a violent event. (Another function is recovery following a violent event, including assistance and support for those directly affected and rebuilding trust and a sense of security in the wider community. Recovery may not be strictly speaking a part of threat management, since it comes after the danger has passed. But it will typically involve the same campus institutions — principally public safety, student affairs, and medical/mental health counseling services — and it is connected to managing threats at least in the sense that measures to improve safety and prevent future incidents have an obvious role in the recovery process.)

An effective threat assessment program, as Deisinger and other speakers noted, requires not just the efforts of an assessment team but awareness, acceptance, communication and cooperation from the rest of the university community. A threat can’t be investigated, evaluated and managed until the team knows about it. That means that faculty (including adjunct faculty, who may be harder to reach and enlist), students, residential staff, public safety officers and other campus staff need to know how and where to inform the right people about a troubling or potentially troubling situation. It also means that administrators who make university policy and the staff that carries it out have to earn and keep the community’s trust. “The people who really know who’s dangerous are peers,” Richard J. Eichler, executive director of health services at Columbia, pointed out. “They won’t come forward if they don’t think the school will act humanely and rationally and with judiciousness, wisdom and compassion.”

Beyond understanding their responsibility to report and how to do it, Deisinger commented, people also need to believe that their reports are wanted and that something will be done about them. They also need to know what to report. Deisinger offered this list:

– Persons at risk of causing harm to others or harm to self

– Persons who demonstrate inability to take care of themselves

– Serious mental health concerns

– Substance abuse

– Behavior that is significantly disruptive to the learning, living, or working environment


Roger Depue, former chief of the FBI’s Behavioral Science Unit and the closing speaker at the Columbia meeting, told the group that relevant information can come from many directions — school records, teacher observations, student or peer information, police reports, counselor concerns, commitment evaluations, and parents’ observations. Depue was a member of the Virginia Tech Review Panel and author of an appendix to the panel’s report, titled “Red Flags, Warning Signs and Indicators,” which said, in part:

A single warning sign by itself usually does not warrant overt action by a threat assessment specialist. It should, however, attract the attention of an assessor who has been sensitized to look for other possible warning signs. If additional warning signs are present then more fact-finding is warranted to determine if there is a likelihood of danger. Some warning signs carry more weight than others. For instance, a fascination with, and possession of, firearms are more significant than being a loner, because possession of firearms gives one the capacity to carry out an attack. But if a person simply possesses firearms and has no other warning signs, it is unlikely that he represents a significant risk of danger. When a cluster of indicators is present then the risk becomes more serious. Thus, a person who possesses firearms, is a loner, shows an interest in past shooting situations, writes stories about homicide and suicide, exhibits aberrant behavior, has talked about retribution against others, and has a history of mental illness and refuses counseling would obviously be considered a significant risk of becoming dangerous to himself or others.


When a threat assessment team becomes aware of someone showing that or a similar list of warning signs, Depue wrote, it could respond by meeting with the student and developing a treatment plan with conditions for remaining in school (or, alternatively, suspending the student until he or she has been treated and doctors indicate the student is not a safety risk). Other possible actions could include contacting parents or guardians; talking with roommates and instructors; requesting permission to receive medical and educational records; and checking to find out about any past restraining orders or encounters with police, especially if they involved anger, stalking, making threats, or using or threatening with a weapon.3

Violence prevention and threat management cannot be reduced to simple formulas or knee-jerk responses. Profiling someone as dangerous from a checklist of factors, or automatic suspension or expulsion under mindless “zero tolerance” policies, is not the way to safer campuses.

“All threats and all threateners are not equal,” the FBI’s National Center for the Analysis of Violent Crime declared in a 2000 report that also warned: “It is especially important that a school not deal with threats by simply kicking the problem out the door. Expelling or suspending a student for making a threat must not be a substitute for careful threat assessment and a considered, consistent policy of intervention. Disciplinary action alone, unaccompanied by any effort to evaluate the threat or the student’s intent, may actually exacerbate the danger– for example, if a student feels unfairly or arbitrarily treated and becomes even angrier and more bent on carrying out a violent act.”4 Virginia Tech’s Gene Deisinger echoed that warning, cautioning against over-reliance on “fear-driven responses” and disciplinary procedures such as suspension, expulsion or termination.  ”Never equate separation with safety,” he concluded.

Different campuses may have different views on how to structure a violence prevention system, what to call it, and what its scope should be. Some schools will avoid the term “threat assessment” because that might suggest that the campus is a dangerous place, and because they define safety and wellbeing more broadly than just protecting their community against possible violent acts. Terms such as “At-Risk Student Support” put a different lens on troubled students or employees, too: not just as threats to others, but as people in need of help.

A system aimed at detecting and dealing with a wider range of emotional problems may have a somewhat different makeup than one that focuses narrowly on potential violence. But in general, both will involve representatives of the university administration, campus police, counseling service and student or faculty/staff assistance programs, the dean of students, the residence office, and the legal department. As well as having different perspectives, those offices may not always completely share the same goals, either. “Administrators want to get rid of the problem, health professionals want to solve the problem, public safety wants to safeguard the campus,” one participant said wryly. Better and more systematic communication will not guarantee unanimity on all policies or procedures or decisions, but as most who attended the Columbia conference appeared to agree, it can make threat assessment (under whatever name) and threat management an effective method for making a campus a safer place for all members of the community.

A Broader View


The way to prevent violence, observes Victor Schwartz, dean of students at Yeshiva University, is not to try to design a way to prevent violence, but to provide good, comprehensive health services for all medical and emotional problems.

“You can’t focus on a specific problem,” Schwartz said in a conversation after the Columbia meeting. “You do it by forming a good medical community and good medical care, picking out the kids who are anxious, depressed, involved in substance abuse.” An effective health and counseling service that is known and trusted by the community will offer the best chance to identify troubled and potentially dangerous people before they become an active threat to themselves or others. “You want to head them off at the pass by providing good treatment, social support,” Schwartz explained. “You’re never going to catch everybody, but providing good networks of care, good community resources, well publicized and well integrated — that puts you in the best position.”

Columbia’s Paul Appelbaum expressed the same view. Seeking to target counseling services only at prevention of severe violence, he said, is likely to be a fruitless approach, because “identifying potentially violent students and intervening with them effectively are very difficult tasks. We are more likely to be effective in reducing violence in campus, violence toward self and toward others, if we focus on providing needed mental health treatment to troubled students.”

Adequate mental health services can be the “the least problematic … the easiest fix” for a range of legal and ethical problems, including those concerning confidentiality and information sharing, Appelbaum commented in his presentation at the Columbia conference. Strengthening those services is thus a concrete policy alternative for university administrators concerned about safety on their campuses. But in an era of financial stress and tight budgets, that approach has not always been adopted, despite heightened awareness and concern after the Virginia Tech and Northern Illinois University shootings. Mental health programs have been expanded on some campuses but demand has also risen, and access to counseling may actually be declining for many students, as utilization grows but funding does not keep pace. (Paradoxically, one possible reason for greater pressure on mental health service at the college level, Appelbaum pointed out, may be that there is more and better treatment of emotional problems among elementary and high school age students. As a result, youngsters who in the past would have been too troubled to finish high school now succeed and make it into college.)

Colleges have not been standing still on this problem. A 2008 report in the Chronicle of Higher Education noted that “over all, support for students with a range of disabilities and emotional issues has grown precipitously over the last 15 years,” but despite those efforts, colleges still “cannot keep up with the rising demand for mental health services.”

In recent surveys by the Association for University and College Counseling Center Directors and the American College Counseling Association, the heads of counseling programs were almost unanimous in reporting that “the number of students with significant psychological problems is a growing concern” on their campuses. Large majorities felt that the population of seriously troubled students has been increasing. (Those trends appeared to be even stronger in larger institutions.) In the comment section of AUCCCD’s 2008 survey report, one director noted: “The psychiatric epidemiological data on persons of traditional college age in the U.S. make it clear that the existing need far outstrips the available treatment resources, on and off campus.” Another wrote: “Somehow the leadership needs to better influence administrators and legislators when it comes to staffing. Tragedies lead to anxiety which leads to finger-pointing but everyone overlooks how short-staffed so many of us are. What is truly amazing is how many tragedies we avoid with the staff sizes we have!”

Two-thirds of respondents in the ACCA survey expressed concern about “the growing demand for services without an appropriate increase in resources” and three-quarters said that an “increase of students with severe psychological problems” is creating administrative issues for their centers. Fewer than a third of ACCA’s respondents, however, reported any increase in counseling staff or psychiatric consulting hours.

With stretched resources, university and college counseling centers typically provide relatively short-term care — on average, five or six visits, after which students needing more extended treatment (which would presumably include those with the most serious problems) are often referred to off-campus providers. Those who need psychiatric care in many cases have to find it off-campus from the start, since nearly one-third of college and university centers do not offer any psychiatric service at all. Even where psychiatrists are available, fully half of center directors reported they could not offer enough psychiatric hours to meet the need. For students who have to get treatment beyond what is available on their campuses, a lack of health insurance coverage may shut them off from the help they need, since, as the Chronicle of Higher Education report pointed out, “many types of counseling and medication that were covered even a few years ago are no longer affordable for students.”5

Faced with those pressures, counseling centers on many campuses will need to become stronger advocates for more adequate and better care, and administrators will have to become more aware that better health services will make their schools safer. The need goes beyond just offering help for troubled students. A broader effort is also needed to develop a healthier campus culture with stronger communal bonds and formal and informal mutual care and support among all students, staff and faculty. When people care about each other and feel their institution cares about them, the odds of detecting someone who is emotionally disturbed and intervening before a tragedy occurs become much better. Frank Ochberg, an eminent psychiatrist and a key creator of the field of critical incident analysis, put it this way: “We’re trying to have healthy campuses, and we’ve learned that with respect to violent outcomes and loss of life, the healthiest campus is one in which kids treat each other in a decent way, and warning signs of destructive behavior are brought to the right resources, to individuals who are good at therapy and good at mentoring.”

Counseling centers cannot change campus culture by themselves, but they can be a catalyst. As Yeshiva University’s Victor Schwartz noted, counselors and student affairs staff members can help teachers and academic advisers “to think beyond students’ academic work, think about them in a more human, holistic way.” To get that message out, he added, he and the head of the university’s counseling service “meet with academic advisers couple of times a year, they hear how we are thinking about cases, trying to put together a picture…. We are constantly speaking with deans. You have to get to faculty meetings, you really have to be in people’s faces, continuously having these conversations. You are modeling an approach that is more holistic, based on looking at things therapeutically. It requires work and commitment.”

As several participants in the Columbia conference pointed out, special outreach efforts will be needed to draw in groups that may not be strongly connected to campus life, such as adjunct faculty and commuter and continuing education students. Communication and education must be designed to reach “people who are not experts” in behavioral science or public safety, Schwartz commented, so that faculty members and students will be more attuned to the needs of troubled students and more aware of the resources available. Outreach will also have to be continuous, since the population of a college or university community is always changing. And, as Fordham University’s Christopher Rodgers told the meeting, there is a constant need for better and more realistic training and for concrete planning on who will respond and how when someone is identified as a potential risk to others’ or his own safety.

No system can guarantee absolute security. Human nature, mental illness and circumstances are simply not predictable enough. Campuses, like the rest of human society, have to live with a certain amount of risk. An additional ambiguity in conceiving and designing prevention strategies is that their effectiveness can never be definitively measured, because there is no way to identify or count or analyze incidents that don’t happen. When prevention fails, of course, the failure is obvious. Sometimes, looking back at a violent tragedy can pinpoint specific precautions that could and should have been taken to keep it from happening. But more often, trying to find a particular barn door that was left unlocked is not the most useful response. A better approach is to build a better barn. That means a prevention plan that guards against common risks, rather than the sensational ones that are much less likely to occur; that relies on a network of collaboration and clear communication, and that contributes to campus-wide awareness and a sense of community, trust, and meaningful human connection.



1 Report of the Virginia Tech Review Panel, Summary of Key Findings, p. 2, and Chapter IV, p. 52.


2 Balancing Student Privacy and School Safety: A Guide to the Family Educational Rights and Privacy Act for Colleges and Universities. U.S. Department of Education, October 2007.

3 Report of the Virginia Tech Review Panel, Appendix M, p. M2-M4.

4 The School Shooter: A Threat Assessment Perspective. The Critical Incident Response Group (CIRG), National Center for the Analysis of Violent Crime (NCAVC), FBI Academy, Quantico, Virginia, 2000; p. 5, 26.

5 See Robert Rando & Victor Barr, Annual Survey, 2008; Robert Rando, Victor Barr, Bryan Krylowicz, & Evelyn Winfield, Annual Survey, 2009, (c) 2009 and 2010, The Association for University and College Counseling Center Directors; and Robert P. Gallagher, National Survey of Counseling Center Directors, 2009, sponsored by the American College Counseling Association, published by International Association of  Counseling Services, Inc., Monograph Series No. 8R; Elizabeth F. Farrell, “Counseling Centers Lack Resources to Help Troubled Students,” Chronicle of Higher Education, Feb. 29, 2008.