… Two separate events spurred me to once again write about this topic. First, the Governor of California signed a bill into law in July of this year that, effective January 1, 2013, deletes the phrase “duty to warn” from the California immunity statute, Section 43.92 of the Civil Code. This law establishes immunity from liability (a “safe harbor”) for a psychotherapist who takes specified actions when the patient communicates to the therapist a serious threat of physical violence against a reasonably identifiable victim. The other event that spurred me to write this article was the shooting in the Aurora, Colorado movie theatre that resulted in the killing and wounding of scores of people. From the moment I heard about the shooting, I wondered whether the shooter was seeing a psychotherapist during the period of time preceding the shooting and what the psychotherapist might have known about the patient’s dangerousness. These two events, occurring close in time, made me once again focus on this thorny, nuanced, and critically important topic – important not only to the mental health community, but to society as well.
In each state, there are three central questions in this area of the law that must be asked and three questions that must be answered. First – when does the duty of a psychotherapist to society, or to an identifiable other, arise with respect to a therapist’s treatment of a patient who threatens violence, if at all. Stated otherwise, when is the duty, if any, triggered? Second – what is the precise duty, if one exists? Third – under what circumstances, if any, should the therapist or counselor be immune from liability in dangerous patient situations, both when certain actions are taken by the therapist and when no action is taken? While the answers to these questions vary from state to state, lawmakers (and others) in some states may want to re-visit these questions (and perhaps develop new answers) in the wake of the shooting in Colorado. This article focuses on the first of these questions, which deals with when the duty is “triggered,” or in the case of states that do not impose a duty, whether a duty should be imposed under specified circumstances.
What is, or would be, good public policy with respect to this aspect of the dangerous patient issue is a matter of disparate opinion. In some states where a duty exists, the duty is only triggered when the patient communicates to the therapist a serious threat of physical violence against an identifiable other. I would argue that as a matter of sound public policy, the duty should be triggered, if there is a duty, when the therapist determines that the patient poses a serious danger of physical violence to a reasonably identifiable victim or victims. This latter trigger of the duty is consistent with the famed Tarasoff v. Regents of the University of California decision of the California Supreme Court (1976), which strikes a balance between the need for confidentiality in therapist-patient relationships and the need for exceptions to be made when physical violence against a readily identifiable victim is imminent.
It is possible that in light of the Colorado case, some states may change their laws to, among other things, create a duty to act where none now exists. How far legislatures go in invading confidentiality remains to be seen. While there should be no duty to predict violence, it is my belief that once the therapist has in fact determined that physical violence against another is imminent he or she should be required to act. While this is an invasion into confidentiality, it is a reasonable one. If a patient telephoned his therapist and threatened imminent and serious physical violence against the therapist, the duty of confidentiality must and should give way. Cancelling the next appointment is not sufficient! The therapist might want, at a minimum, to inform the police of the threat and of the existing relationship of therapist-patient.
A California appellate court has interpreted the California immunity statute, which some interpret as saying that the duty to protect an intended victim is only triggered when a threat is communicated by the patient to the therapist. The court held that the communication does not have to come directly from the patient. In that case, the father of the patient told the therapist about his son’s (the patient’s) threat of violence –communicated to the father. The court held that this communication from the father could be considered a communication from the patient since the father’s communication to the therapist was properly considered to be a part of the therapy and for the benefit of the son’s treatment. The court also considered the spirit and purpose of the immunity statute in making its determination. Further, I have long thought that a court might someday rule that the patient’s communication of a threat can even be delivered nonverbally, and that this nonverbal action or behavior could trigger a duty for the therapist to take some action to protect another. All it takes for that to happen is for the right kind of case to come along.
For example, suppose that a long-term patient with a history of violence, who is angry at his former supervisor at work for the recent firing and the humiliation he suffered, unilaterally and suddenly terminates treatment by leaving a cryptic message with the therapist’s answering service. Some therapists might be led to conclude that the termination was an indication of likely imminent violence (when taken together with what had transpired before, including the patient’s history of violence). In the Aurora shooting, some reports seem to suggest that the psychiatrist notified a threat assessment unit at the school (we do not yet know what the content of the notification was), but that no further action was taken because the psychiatrist and the school found out that the alleged shooter was dropping out of school. Could the act of dropping out of school, instead of triggering no further action, be seen as “the straw that should have broken the camel’s back” – thus triggering a duty to protect potential victims? This question might eventually be asked by Colorado legislators if they consider making any changes to Colorado law. The facts in that case present the additional question of whether reasonably identifiable patients must be known, or whether it is enough that the patient presents a danger to the public in general.
Lest you think that I undervalue the need for confidentiality – to the contrary. I merely believe there are limits. It is useful to at this point quote some of the language from the Court’s decision in Tarasoff, which I think expresses this issue squarely and appropriately. “We realize that the open and confidential character of psychotherapeutic dialogue encourages patients to express threats of violence, few of which are ever executed. Certainly a therapist should not be encouraged routinely to reveal such threat; such disclosures could seriously disrupt the patient’s relationship with his therapist and with the persons threatened. To the contrary, the therapist’s obligations to his patient require that he not disclose a confidence unless such disclosure is necessary to avert danger to others, and even then that he do so discreetly, in a fashion that would preserve the privacy of his patient to the fullest extent compatible with the prevention of the threatened danger.”
In summary then, on the issue of when the duty, if there is one, should be triggered, thought should be given to the idea that limiting the trigger to an articulated threat, rather than to an assessment by the practitioner based upon all of the facts and circumstances extant, including the history of the patient, may be too limiting and unnecessarily protective of practitioners. If the duty to take action to prevent the threatened violence, which may involve breaking confidentiality, is triggered when a therapist determines imminent dangerousness (understanding that there is no duty to predict dangerousness), this leaves the practitioner in a relatively good position.
With respect to the bill in California that will become law shortly, there was no substantive change made to the immunity statute. The bill simply removed the terms “duty to warn” and “warn of” from the statute, so that when the statute refers to the duty, it correctly calls it a duty to protect rather than a duty to warn and protect. There is not now, nor has there been in the last thirty plus years, a duty to warn. There are reasons why some have gotten this wrong, including the courts, but I need not explore those reasons here and now. The California immunity statute makes clear that the duty to protect shall be considered discharged and the therapist will be immune from liability if the therapist makes reasonable efforts to communicate the threat to the intended victim(s) and to law enforcement (typically the police or sheriff).
The duty to protect, however, need not be discharged in the manner specified in this statute. If one desires immunity from liability, the statute must, as above-stated, be followed. The therapist is free, however, to take other action, such as, hospitalizing the patient, referring for medication and psychiatric evaluation, or to taking whatever other steps are reasonably necessary to prevent the threatened harm. Although the therapist in such an instance will not have immunity from liability, it may well be that the therapist acted reasonably and that no liability will be found by judge or jury.