Abstract
In 2013, legislation went into effect clarifying that the Tarasoff duty in California is now unambiguously solely a duty to protect. Warning the potential victim and the police is not a requirement, but a clinician can obtain immunity from liability by using this safe harbor. In situations in which a therapist believes warning might exacerbate the patient's risk, however, alternative protective actions can satisfy the duty to protect. For a clinician to be found liable, those alternative actions would have to be proven negligent. This flexibility can sometimes be crucial in protecting potential victims and thereby, indirectly, patients from the consequences of dangerous action. Explaining the reasoning for the action chosen should obviate any significant liability risk of doing the right thing, even without immunity. Legislation was enacted in 2007 as an attempt to clarify the requirement, but the revised immunity statute at the time retained the phrase duty to warn and protect, which perpetuated the now-eliminated confusion. Correctly understanding the California law is important to avoid having the restored flexibility eroded again by belief in a nonexistent duty to warn. The Tarasoff duty originated in California, but since many other states later established similar duties, the developments in California may have national implications.
This article is intended to clarify and update information about the current state of California law regarding therapists' duties after a determination that their patients are at risk for acting dangerously; to describe the history of the Tarasoff duty in California with a historical account of related judicial decisions and statutes, including recent changes; to explain how a nonexistent duty to warn influenced the law's trajectory, despite being the law for only two years in the 1970s and again recently for several years after appellate court decisions; to provide clinically relevant examples in practice that were used to influence a change the law; and to discuss current implications in California and elsewhere.
In California, the state where Tarasoff v. Regents of University of California1,2 itself was decided, a duty to a potential victim was found based on the special relationship between doctor and patient. The Tarasoff duty after the 1976 ruling2 was and is now again solely a duty to protect. A duty to warn existed only from 1974 to 19762 and more recently from 2004 to 2006. In the latest definitive clarification effective January 1, 2013,3 all references to a therapist's duty to warn were completely removed from the relevant immunity statute. An earlier revision, in 2007, did not accomplish all that was intended.4 Those events were described in The Journal in 2006.5 At that time, to ensure passage, language referring to a duty to warn and protect was retained in the 2006 revision effective 2007 (hereafter, the 2007 revision), despite clarification in the statute itself and the subsequent jury instructions. The continued reference to a duty to warn and protect contributed to the persistent erroneous belief by some that there still was a duty to warn in California. Any remaining legitimate confusion was eliminated and clarified in the most recent revision.3 The new legislative revision did not change the meaning of the 2007 revision that already had removed any duty to warn. It simply removed any ambiguity about the meaning of the revision. The duty is to protect, with immunity obtained if the therapist chooses to warn. However, there is no duty to warn.
The duty already had been clarified further in the California Judicial Council revised jury instructions in 20076 that followed the 2007 revision to combat two problematic appellate court decisions that had for a brief period resurrected a duty to warn. When a patient makes a credible specific threat toward a potential identifiable victim, warning that victim may or may not be a protective course of action. Warning remains relevant in situations where a duty to protect is triggered. Immunity from liability ensues if the therapist chooses to take reasonable steps to warn a potential victim and notify the police. This procedure thereby encourages, but does not mandate, warning. If the therapist thinks warning would increase the risk and instead chooses an alternative protective action, the therapist's actions must be proven negligent for the therapist to incur liability, much as in other areas of medical malpractice.
Historical Evolution of Tarasoff Duties
In 1974, the California Supreme Court established an unprecedented duty to warn based on the special relationship between the therapist and the patient (as well as potential police liability). Tarasoff I1 was troubling to therapists in California and around the country for its then unprecedented requirement to violate patient confidentiality. At that time, patient-therapist confidentiality was more inviolable than it is now. Any violation to protect the public was optional on the part of the therapist, much as it now is for attorneys in California. Police were likewise worried about far-reaching liability for releasing potentially violent individuals. As a result, the California Supreme Court reheard the case (Tarasoff II),2 and removed police liability and replaced the duty to warn with a duty to protect. Warning then as well as now was merely one method of satisfying the duty to protect, but was not required or necessary and certainly was not the only way to satisfy the duty to protect. The claimed advantage of choosing to warn was its debatable perception of being less of an intrusion on the patient than an involuntary hospitalization that deprives liberty. No immunity in California was provided by any action at that earlier time.
Although these rulings applied only in California, the effects reverberated nationally. Many but not all jurisdictions later developed similar duties through case law and legislation. A provision in the Tarasoff II2 decision in 1976 held that a therapist could be liable if he “should have” known that a patient was dangerous before the patient engaged in a harmful act. This ruling led to what mental health professional organizations saw as unpredictable and therefore unreasonable therapist liability. If the patient did something dangerous, it was easy retrospectively to think that a therapist should have known. In some states with Tarasoff-type obligations, therapists had been found liable even for injuries to victims of car accidents caused by former patients who had been evaluated months earlier by the therapist and who were under the influence of alcohol or drugs at the time of the accident.7 Liability was found, despite the therapist's having no way to stop the drug use or prevent the accident. Lengthy hospitalizations were not legally an option, nor was it possible to predict the accident. Some jurisdictions expected therapists to foresee all dangerous situations and to protect even unidentifiable victims. As a result, California and many other jurisdictions passed immunity statutes specifying the situations creating a duty to protect and a means to obtain immunity. The 1986 California immunity statute8 granted therapists immunity from potential future liability if they made reasonable attempts to warn a potential victim and to notify the police. Unlike some other states, in California, involuntary hospitalization did not and does not confer immunity.
As in many states, the statute limited the duty to protect and the potential liability to situations in which the patient communicated to the therapist a serious threat to an identifiable victim. The words of that original California immunity statute were ambiguous, though, and made reference to a “duty to warn and protect.”8 The language most likely was intended to be inclusive, since many erroneously continued to refer to a duty to warn, even though such a duty had not been the law for many years. We could find no evidence that anybody intended the immunity statute to create a new duty to warn or to reject the 1976 California Supreme Court Tarasoff II2 decision and revert to the earlier 1974 Tarasoff I1 criteria.
Nearly two decades after passage of the original immunity statute, the duty to warn was resurrected after California's 2003 simplified civil jury instructions (California Civil Instructions; CACI) interpreted the ambiguous 1986 immunity statute to have also created a new duty to warn, most likely because the “warn” portion of the duty to warn and protect was interpreted to refer to a duty that could be satisfied only by warning.6 Two appellate court decisions in 2004, Ewing v. Northridge Hospital Medical Center9 and Ewing v. Goldstein,10 instigated the return of a duty to warn for several years in California. The intent of the simplified jury instructions was not to change them, but the California Judicial Council (which publishes CACI) most likely misinterpreted the statute, because of its ambiguity, to have created a new duty to warn. The appellate courts in Ewing v. Goldstein10 went even further and interpreted the new duty to warn so rigidly that, if a serious threat to an identifiable victim was communicated to the therapist by the patient or the court-determined equivalency of an immediate family member, the duty to warn became automatic, as did the therapist's liability in the event of no warning. Lay jurors, according to the Ewing v. Goldstein decision, with no requirement for expert testimony, could be expected to determine whether a serious threat to an identifiable victim was communicated to the therapist, whether the therapist considered it credible, and whether the therapist had warned. If there was no warning once the described duty was triggered, nothing else was necessary for the therapist to be found liable if there was harm to a victim. Alternative protective actions apparently were irrelevant, since the absence of a warning alone would create liability. The decision created serious problems for responsible therapists.
First Statutory Revision to Correct the Problem
As a result of these court rulings, the therapist might not even have an opportunity in court to explain the reasons for not warning or for taking alternative and more protective actions, and if he was allowed to testify, his explanation seemingly would not be relevant. The only question was whether the therapist had warned. If not, and if a threat was acted on with damage to a threatened victim, automatic liability followed. Although there are no statistics, it seems possible that, during this period, some therapists, fearing automatic liability, may have given counterproductive warnings that could have exacerbated the danger. The warning requirement in many instances did not protect victims. Furthermore, these interpretations created unreasonable liability for conscientious therapists who, on occasion, thought that warning would increase the risk to others and, therefore, to protect potential victims, did not warn. If a patient later harmed a threatened victim, the therapist would be liable automatically. Meanwhile, some therapists probably augmented the danger posed by the situation by giving irresponsible warnings that exacerbated conflict, out of realistic fears of liability if they did not warn and somebody was harmed. This serious dilemma for conscientious therapists existed for several years before the law was changed, effective in 2007. The revision restored the duty to protect and no longer required warning. This legislative change was intended to overturn the 2004 Ewing precedents.
The first author's work with the California Association of Marriage and Family Therapists (CAMFT) and consultation to the California Judicial Council in his role at the time as chair of the Judicial Action Committee of the California Psychiatric Association (CPA) has helped influence legislation since 2006, to effect changes in the immunity statute and corresponding revision of the jury instructions. Those active in this process hoped that the initial changes would be sufficient to remove the erroneous impression that the original immunity statute had created a new duty to warn. Political considerations, however, led to the retention of the language duty to warn and protect in the immunity statute. An influential legislative individual misinterpreted the original proposal to mean advocating unusual unprofessional actions, such as slashing a patient's tires to prevent the patient from posing a danger and said, “The patient could then just take a bus.”
As a result of this process, it was necessary to retain the phrase “duty to warn and protect” in the 2007 revision,4 despite language in the statute that warning was merely a way to satisfy the duty to protect. It was hoped that misinterpretation would be avoided by revision of jury instructions consistent with the legislative intent.
Revised Jury Instructions
The California Judicial Council revised the jury instructions in response to the revision to clarify the intent to remove any duty to warn. The most significant changes to improve the jury instructions were made in 2007, but some small additional changes were made in 2013 to simplify some wording and clarify that reasonable efforts to warn the victim and the police were sufficient to get immunity consistent with the revised statute without needing to actually warn them.11
Instruction 503A is to be read in actions for professional negligence against a psychotherapist for failure to protect a victim from a patient's act of violence, after the patient communicated to the therapist a serious threat to the victim. It clarifies that the usual malpractice standard of negligence is applicable for liability if there are failures to take reasonable measures to protect a victim.
If the therapist claims immunity from liability because he or she made reasonable attempts to warn the intended victim and notify the police, but such a claim is disputed as a factual matter, then instruction 503B is also given. It is an affirmative defense offered by a defendant therapist that, if proven, would confer immunity.
Jury instructions, though, are not binding law and are relevant only at trial. Clinicians rarely refer to jury instructions. So, much confusion seemed to persist and other action became necessary.
Case Examples
The following case examples illustrate problems that arose during the brief resurgence of the duty to warn. They were presented to the California Judicial Council to illustrate the importance of the adopted legislation to overturn those decisions and the need to revise the relevant simplified jury instructions to restore clinical flexibility to permit decisions in the best interest of patients and society. These specific examples helped to effect changes to the first revision of the immunity statute, then to the revised simplified jury instructions, and finally to the recent California statutory amendment removing any remaining ambiguity that there is no duty to warn but only a duty to protect.
Case 1: Warning May Exacerbate a Dangerous Scenario
A patient entered the hospital after saying that she wanted to kill her father. The father was being released from prison after completing his sentence for killing the patient's mother. He had recently threatened her because he wanted the mother's monetary inheritance, which had gone to the daughter as a result of her mother's death at the hands of the father. On admission, the patient said that she wanted to kill him, but after a day on the inpatient unit, she calmed down and credibly said she was simply angry and started thinking of constructive alternatives to protect herself from her father. Under the then-recent Ewing9,10 court interpretations of Tarasoff liability, there was no flexibility and no option but to warn the father of the threat, despite his history of murder and his recent threats. If she was released and killed her father, there would be automatic liability for the hospital physicians for not warning, regardless of their rationale. Liability was automatic without any opportunity afforded for the physician to explain the reasoning for not warning. Despite the liability risk, the involved clinicians in this case thought it much more likely that the father would kill the daughter than the reverse and that warning him would actually make the situation more dangerous. It seemed irresponsible and even unethical to warn the father, despite the liability that failure to warn created if the threat was acted on. The clinicians were disturbed that they could face automatic liability for possible damages for doing what seemed most ethical and clinically appropriate. Other similar examples include warning the perpetrator in abusive situations of a threat by the victim, or warning feuding parties in custody disputes of fleeting threats by one against the other.
Case 2: Warning May Not Be Feasible or Warranted at the Time of Evaluation
An admitting clinician who had not provided regular inpatient care for a threatening individual could, during this brief interpretation of California law, be in a bind on such an individual's discharge, even if a warning was not feasible or not clearly warranted on hospital admission. As in Ewing v. Northridge Hospital,9 if the admitting clinician gave no warning to a threatened victim, he would be liable if the patient was discharged by the inpatient team and the patient killed or harmed this victim. Retrospectively, hospitalization might itself provide evidence that the admitting clinician thought the patient was dangerous. This example was pertinent to resident physicians in teaching hospitals if they made no reasonable attempts to warn a potential victim in the middle of the night; specifically, they would have been liable for not warning if there was a dangerous action resulting in damages and, based on later clinical assessment, no other clinician had warned. If the patient was admitted, thus protecting the intended victim, but was later discharged based on other clinicians' assessments without warning, the admitting clinician would be liable for not warning under this Ewing decision, regardless of other protective measures.
Case 3: Dangerousness Assessed Before Completion of a Full Evaluation
Under Ewing, therapists who thought a patient credibly dangerous at any point during an interview seemingly were subject to a rigid duty to warn, despite changes or evolution in their clinical judgment over the course of the evaluation.9,10 For example, if a patient made a threat but reneged within minutes and the therapist believed the patient dangerous at the moment of the threat, such conduct would have triggered the rigid duty to warn, as the immunity statute then was interpreted. Neither the patient's change of mind nor a reassessment would have obviated the duty. Therefore, it was important to revise the immunity statute to negate the interpretations of it at that time.
Current Revised Immunity Statute
Even after the jury instructions were revised effective 2007 to reflect only a duty to protect, it is likely that many jurists, practitioners, and clinicians did not look past the statute that still made reference in 2007 to a duty to warn and protect.5 Many even seemed unaware of the change. Some may have seen a duty to warn and protect in the revised immunity statute and continued to believe erroneously that there was a duty to warn. Few such cases go to trial where the jury instructions become relevant and noticeable, and thus continued misinterpretation of the statute by clinicians remained a major problem and risked undermining the revisions again, as in 1976. Confusion by some with the mandatory reporting required of mental health professionals in situations of suspected child or elder abuse may have led and may continue to lead to the mistaken belief there is an equivalent mandatory duty to warn, even after the latest 2013 clarification. Some consulting attorneys, to whom clinicians may entirely defer for decision-making, may have compounded the confusion with a focus on risk aversion alone and the highest level of liability protection conferred by warning, ignoring the patient and societal welfare that clinicians should consider, because regardless of adverse consequences, warning would confer immunity. The duty to warn was in effect recently only from 2004 to 2006, but many erroneously continue to refer to a duty to warn, much as they did after Tarasoff II.2
As a result of the persistence of a belief in an already nonexistent duty to warn, it became necessary recently to modify the California immunity statute again, not to change the duty but to clarify it and remove all remaining ambiguity. The continued erroneous reference by many to the already nonexistent duty to warn made clear that any reference at all to a duty to warn should be removed from the immunity statute. That goal was unambiguously accomplished in the 2012 revised statute that took effect January 1, 2013.3 All references to any duty to warn were completely removed from the statute. Henceforth, if a therapist chooses not to warn, but instead pursues an alternative course of action for clinical and ethical reasons, such actions must be proven negligent to find legal liability, as in other areas of malpractice, just as the revised jury instructions make clear.6
With the recently amended immunity statute3 (see Appendix), all ambiguity about the current status of the Tarasoff duty is removed. The duty has been clarified, but is unchanged from the previous potentially ambiguous 2007 revision.4 There no longer is any legitimate reason to believe in California in a non-existent duty to warn. It is essential that clinicians and their advisors become aware of the unambiguous status of the current law.
Discussion and Opinion
In our opinion, the current California framework provides a good balance, permitting and even encouraging violations of confidentiality if the therapist believes there is a serious imminent threat. To maintain immunity, the therapist can deliver warnings to potential victims and make reports to the police and others as well, if such actions seem necessary for protection from violence; but there are alternative options.
The short-lived former duty to warn in our opinion did not necessarily protect victims. It is rare that the police or the victim can stop the patient, unless the patient is in the process of committing a violent act. The victim may be able to avoid the threatening patient, but there are often risks of exacerbating the conflict. Sometimes the potential victim can get a restraining order, but this measure is helpful only if the patient approaches the victim and the police are called. Restraining orders can be inflammatory, and there may not be sufficient time for the police to respond, even if the victim sees that the patient is intent on violence. A therapist working with the patient to diffuse the conflict may be the most effective way to protect the potential victim and to help the patient avoid the serious consequences of committing a violent act. Furthermore, the rigid duty to warn (for the few years that the Ewing decisions9,10 provided the precedent) created unreasonable liability for conscientious therapists who, on occasion, thought that warning would increase the risk to others and therefore did not warn.12 In any event, after 1986, the warning did and still does confer immunity from liability.7
Despite there having been no duty to warn since 1976 (except from 2004 to 2006), references were and still are made to a duty to warn in California and elsewhere when describing the California duty. Perhaps it is because so much attention was given to the original duty-to-warn decision, notable at the time for its unprecedented violation of patient confidentiality.5 Perhaps it is because most clinicians do not keep abreast of legal developments in detail, or perhaps it is because the replacement of the duty to warn with a duty to protect received a fraction of the attention that the earlier duty to warn had received. It may also be that even now some attorneys and risk management consultants think that clinicians care more about immunity and protecting themselves than in doing the best thing to protect potential victims despite the limited liability risk.
The only remaining portion of the Ewing decisions9,10 that is in force is that communication of a serious threat via a close family member is treated as the equivalent of a communication directly by the patient to the therapist and is described in a footnote in the revised jury instructions.6 In reality, warnings can be useless in conferring protection of the victim, because warnings do not address the cause of the threat. Even worse, as mentioned earlier, warnings can exacerbate the conflict and increase the risk of dangerous action.
Although therapists usually will want to warn and report to the police to obtain immunity and protect themselves, under current law, therapists again have the freedom to decide not to warn and instead to engage in alternative, more protective action. Although concerns have been expressed that anything but mandated warnings can increase liability, the California framework addresses that concern by giving the safe harbor of immunity to anyone who wants it and chooses to warn.12 Deciding on an alternative action is an option when a therapist thinks it is more important to avoid exacerbating the problem and instead to accomplish something more protective.
Potential Effect Beyond California
Although the revised duty-to-protect statute is not applicable outside of California, the unambiguous removal of a duty to warn could have implications in other jurisdictions. California first created a duty to warn and developed the reasoning behind it in 1974.1 In 1976, the duty was changed to a duty to protect, with warning being only a way to satisfy the duty to protect.2 Despite the revised California Supreme Court opinion, many if not most in California and elsewhere retained a mistaken belief in the presence of a duty to warn. The Tarasoff duty frequently was erroneously called a duty to warn. Although jurisdictions around the country differed in whether there was a duty to warn or protect and if so what the duty was, many followed California's laws. There is little evidence that those jurisdictions that adopted a duty to warn knew that California had eliminated it, and these jurisdictions adopted a duty to warn, knowingly rejecting the duty to protect. Instead, as in California itself, it seems most likely that there was a mistaken belief that California had maintained a duty to warn, and these states adopted an analogous duty. There seemed to be an implicit assumption that warning would be protective and never counterproductive. Of course, just because California has now unambiguously corrected its error does not mean that other jurisdictions will as well. In fact, there continues to be some risk that, despite language in the California law that abolishes any ambiguity, many, even in California, are not aware that anything has changed, much as there has been a belief for years that California had maintained the duty to warn long after it was eliminated in 1976.
The fact that the law, at least in California, is now clear should have an impact even outside of California, if the information becomes well known. States that have developed such duties have done so by various legislative or court decisions.13 This article in part is an effort to help publicize the clear determination that the California duty now is solely to protect and not to warn. Since California was the leader in this area, it is hoped that some other jurisdictions will make changes as well for the same reasons as California. Much like the California effort, it would help to emphasize that the change can be more protective of victims, avoiding counterproductive warnings that can inflame situations and increase the risk of a dangerous action. Unless the changes are known by most therapists in California and elsewhere, nothing is likely to change, despite the revised statute, as happened after the 1976 Tarasoff II decision in California, where the incorrect perception that there was still a duty to warn contributed to its temporary actual resurgence for a few years.
Clinical flexibility in California now allows the clinician to take measures to diffuse dangerous situations and does not mandate warnings when it is thought that they may be inflammatory or counterproductive. Forensic psychiatrists in California should be aware of current law so that they do not erroneously claim that warning is the only way to satisfy the duty. If unaware of the changes they could claim that taking other, potentially more protective action is necessarily negligent, absent a warning. Although some other states still have a duty to warn, hopefully the rationale for changing the California law will have impact in other jurisdictions based on the likelihood that clinical flexibility sometimes would be most protective of potential victims. Some case examples strongly suggest this, though we are unaware of any relevant statistical data. Just as the rationale for the Tarasoff duty was used to establish similar duties in other jurisdictions, including a misguided duty to warn, it is to be hoped that knowing and understanding the rationale for removing the duty to warn will help persuade other jurisdictions to replace the duty to warn with a duty to protect.
Conclusion
There is no longer a duty to warn in California. Both warning potential victims and notifying the police provide immunity from liability. However, it is not necessary to obtain immunity to avoid liability. In most circumstances, therapists will want to warn potential victims and the police, to obtain both immunity for the therapist and possible protection for the potential victim. Warning the potential victim and notifying the police, however, will sometimes be useless in actually providing protection to the victim, with other actions being more protective. Although immunity for the therapist is obtained from warning, there may be negligible protective value for the victim, and sometimes the warning may inflame the situation.
Other actions that do not lead to immunity for the therapist may well be more protective. Examples include hospitalization, medication management, or other therapeutic interventions (including reality testing, cognitive restructuring, or supportive therapy). All of these (unlike warning) may well diffuse the danger and can resolve the problem definitively. These protective actions can supplement warning for those therapists desiring the safe harbor against liability of warning the potential victim and the police. It is usually helpful to make efforts to diffuse the conflict that may be the basis for the threat. Ethically and clinically, protective actions taken in addition to or, occasionally, instead of breaching confidentiality, can be the best courses of action. Warning alone rarely eliminates the danger.
Responsible therapists should consider taking a small liability risk and doing something that is actually protective. Although warning may provide legal immunity, a counterproductive warning is not clinically or ethically warranted and, again, in California is no longer mandated. To be found liable for an alternative protective action, a plaintiff would have to prove that the therapist's action was negligent. That standard is the usual one in malpractice liability, and so, just as in other areas of mental health practice and all of medicine, there should not be excessive liability fears for responsible action, even without immunity. Also judges and juries are likely to be more impressed by clinicians trying to do the most protective thing for patients as opposed to merely protecting themselves. In situations where there are irresponsible actions, courts might even develop new theories of liability. That may have happened in the original Tarasoff case itself.
Thorough documentation is advised, to explain the reasons for a decision not to warn and to indicate that those steps were considered. Although not required, it is probably best to explain why a decision was made not to warn. Documentation helps in proactively addressing future questions, should the therapist be wrong and a dangerous action result. As with other potential malpractice situations, contemporaneous notes help by showing the thought process and specific reasons for any decision.
Clinical flexibility can be crucial in diffusing threats, as opposed to reflexive and sometimes counterproductive warnings. It is important for confusion not to result again in loss of this flexibility. To prevent a recurrence of the erosion of flexibility, as happened despite Tarasoff II in 1976, California therapists must be clear that the current duty is only to protect. Warning is the only way to ensure immunity, but is not the only way to avoid liability.
Appendix
Amendment of Section 43.92 of the California Civil Code (effective 2013)
Section 43.92 of the Civil Code is amended to read: 43.92. (a) There shall be no monetary liability on the part of, and no cause of action shall arise against, any person who is a psychotherapist as defined in Section 1010 of the Evidence Code in failing to protect from a patient's threatened violent behavior or failing to predict and protect from a patient's violent behavior except if the patient has communicated to the psychotherapist a serious threat of physical violence against a reasonably identifiable victim or victims.
(b) There shall be no monetary liability on the part of, and no cause of action shall arise against, a psychotherapist who, under the limited circumstances specified in subdivision (a), discharges his or her duty to protect by making reasonable efforts to communicate the threat to the victim or victims and to a law enforcement agency.
(c) It is the intent of the Legislature that the amendments made by the act adding this subdivision only change the name of the duty referenced in this section from a duty to warn and protect to a duty to protect. Nothing in this section shall be construed to be a substantive change, and any duty of a psychotherapist shall not be modified as a result of changing the wording in this section.
(d) It is the intent of the Legislature that a court interpret this section, as amended by the act adding this subdivision, in a manner consistent with the interpretation of this section as it read before January 1, 2013.
Footnotes
Disclosures of financial or other potential conflicts of interest: None.
- © 2014 American Academy of Psychiatry and the Law
