# Carole Kuligoski, Individually and On Behalf of Michael J. Kuligoski, and Mark Kuligoski and James Kuligoski v. Brattleboro Retreat and Northeast Kingdom Human Services

> Supreme Court of Vermont · September 16, 2016 · 203 Vt. 328

URL: https://www.frixlaw.com/law-library/cases/9836153

## Case

- **Full name:** Carole Kuligoski, Individually and on Behalf of Michael J. Kuligoski, and Mark Kuligoski and James Kuligoski v. Brattleboro Retreat and Northeast Kingdom Human Services
- **Court:** Supreme Court of Vermont
- **Decided:** September 16, 2016
- **Citations:** 203 Vt. 328; 156 A.3d 436; 2016 VT 54A; 2016 Vt. LEXIS 106
- **Precedential status:** Published
- **Opinion:** Dissent by Reiber
- **Judges:** Reiber, Dooley, Skoglund, Robinson, Morris, Supr
- **Cited by:** 12 later opinions in the Frix Law Library

## Citator (automated)

- No negative treatment found by the automated citator. That is not the same as a confirmation that the case is good law; read the citing cases.
- Full citator and citing cases: https://www.frixlaw.com/law-library/cases/9836153

## How later opinions describe it (automated extraction)

- explaining that most states balance "the competing concerns by recognizing a relatively narrow duty of care limited to situations where the therapist knows or should know that a patient poses a specific threat to an identified or reasonably identifiable third person"
- noting that "several courts have limited the duty to identifiable victims, or a class of individuals whose injury is foreseeable because of their relationship or proximity to a specifically identifiable victim"

## Opinion text

Reiber, CJ.,
¶ 85. dissenting. Chief Justice Roger Traynor of the California Supreme Court, one of the great common-law innovators in American legal history, nevertheless repeatedly cautioned restraint, or what he called “circumspection,” in the evolution of judicial precedent. “The greatest judges of the common law have proceeded in this way,” he explained, “moving not by fits and starts, but at the pace of a tortoise that explores every inch of the way, steadily making advances though it carries the past on its back.” 17 Unlike a legislature, whose scope of inquiry is unbounded, an appellate court is confined to the record, which in turn is limited by the rules of evidence, and its decisions — unlike statutes — become instantly resistant to change under the rule of stare decisis. Hence the overarching need for judicial humility in the face of our own limited knowledge — for incremental rulings that allow a court “time to advance or retreat” from its forays into the unknown with a minimum of unintended effects and needless shock to those who must “act in reliance upon judicial pronouncements.” 18
¶ 86. The majority abandons this cautious approach with no apparent awareness that it is even doing so. It dresses its decision in the clothes of the “modern,” suggesting that its holding flows from a natural “evolution of the duties articulated in decades of case law” and thus represents no dramatic departure. Ante, ¶¶ 21, *373 23, 38. It embraces these “precedents and modern trends” to define for mental health care providers a new common-law duty. Ante, ¶ 51.
¶ 87. But the argument is a fiction. Science and the law have indeed evolved in the forty years since the California Supreme Court’s seminal decision in Tarasoff v. Regents of University of California that a therapist who “determines, or pursuant to the standards of his profession should determine, that his patient presents a serious danger of violence to another, . . . incurs an obligation to use reasonable care to protect the intended victim against such danger.” 551 P.3d 334 , 340 (Cal. 1976). They have simply not evolved in any way that remotely supports the majority’s decision to expand exponentially the duty owed by a mental health professional to protect third parties in the circumstances presented here. Accordingly, I must respectfully dissent.
¶ 88. The majority observes at the outset that since Tarasoff “several courts have limited the duty to identifiable victims, or a class of individuals whose injury is foreseeable because of their relationship or proximity to a specifically identifiable victim.” Ante, ¶ 36 (emphasis added). Among these, of course, is our own holding in Peck v. Counseling Service of Addison County, Inc. that “a mental health professional who knows or, based upon the standards of the mental health profession, should know that his or her patient poses a serious risk of danger to an identifiable victim has a duty to exercise reasonable care to protect him or her from that danger.” 146 Vt. 61, 68 , 499 A.2d 422, 427 (1985). “However,” the majority continues, “several other courts have held that a duty to warn is owed not only to specifically identified or identifiable victims, but to foreseeable victims or to those whose membership in a particular class . . . places them within a zone of danger.” Ante, ¶ 37. Combined with the suggestion that “Peck was decided thirty years ago, before modem trends in this area,” ante, ¶ 38 (emphasis added), the implication is that the states are now about evenly divided between these camps.
¶ 89. This is decidedly not the case. The voluminous literature canvassing the legal and medical ramifications of Tarasoff over the past four decades agree that the predominant legal response has been to specifically define and limit a mental health provider’s duty to protect third parties, generally requiring a serious threat to a readily identifiable victim. See, e.g., D. Katner, Confidentiality and Juvenile Mental Health Records in Dependency Pro *374 ceedings, 12 Wm. & Mary Bill of Rt. J. 511, 532 (2004) (Although “most jurisdictions now recognize a Tarasoff-type duty, the vast majority . . . limit it to situations in which ... the patient has communicated to the psychotherapist a serious threat of physical violence against a reasonably identifiable victim or victims.” (quotations omitted)); C. Cantu, et al., Bitter Medicine: A Critical Look at the Mental Health Care Provider’s Duty to Warn in Texas, 31 St. Mary’s L.J. 359, 377 (2000) (“The majority of states that have addressed this issue follow the Tarasoff/Thompson rule, which states that when a mental health care provider foresees or should foresee that a patent poses a serious risk of violence to a readily identifiable third person, a duty arises to use reasonable care to protect that individual against the danger.”). 19 See also Fraser v. United States, 674 A.2d 811, 816 (Conn. 1996) (noting that “state courts . . . have overwhelmingly concluded that an unidentifiable victim has no claim in negligence against psychotherapists who were treating the assailant on an outpatient basis”); Eckhardt v. Kirts, 534 N.E.2d 1339, 1344 (Ill. App. Ct. 1989) (observing that, in “determinting] the legal duty of therapists to third persons, numerous courts have concluded that a therapist cannot be held liable for injuries inflicted upon third persons absent specific threats to a readily identifiable victim”). 20
*375 ¶ 90. The reason is readily apparent. Courts and legislatures from Tarasoff onward have recognized the conflicting interests at play in such cases and the freighted consequences however the balance is struck. On one side is the obvious and compelling interest in protecting the public from assault by mental health patients with violent propensities. On the other is the strong countervailing interest in safeguarding the confidential character of psychotherapeutic communications, 21 the inherent difficulty (often underappreciated by those with the luxury of hindsight) of forecasting future dangerousness 22 and the significant societal concern that patients not be unnecessarily hospitalized as a means to avoid liability. 23 See, e.g., Estates of Morgan v. Fairfield Family Counseling Ctr., 1997-Ohio-194 , 673 N.E.2d 1311, 1322 (listing the factors generally considered in determining a therapist’s duty of care as including “the public’s interest in safety from violent assault,” the “difficulty inherent in attempting to forecast whether a patient represents a substantial risk of physical harm to others,” *376 the “goal of placing the mental patient in the least restrictive environment . . . free from unnecessary confinement,” and the “social importance of maintaining the confidential nature of psy-chotherapeutic communications”).
¶ 91. A few states, weighing these countervailing concerns, have determined that public policy simply does not support the imposition of any duty upon a mental health care provider to protect third parties from a potentially violent patient. See Boynton v. Burglass, 590 So. 2d 446, 448 (Fla. Dist. Ct. App. 1991) (rejecting Tarasoff-like duty to warn identified third parties of threats by patient on the ground that it is “neither reasonable nor workable and is potentially fatal to effective patient-therapist relationships”); Thapar v. Zezulka, 994 S.W.2d 635, 640 (Tex. 1999) (declining “to impose a common law duty on mental-health professionals to warn third parties of their patient’s threats”).
¶ 92. Several other courts have taken the opposite tack, broadly defining the therapist’s duty to include any “foreseeable” victim without limitation to specifically identified or identifiable targets of violence. See Lipari v. Sears, Roebuck & Co., 497 F. Supp. 185, 194 (D. Neb. 1980); Naidu v. Laird, 539 A.2d 1064, 1072-73 (Del. 1988); Petersen v. State, 671 P.2d 230, 237 (Wash. 1983); Schuster v. Altenberg, 424 N.W.2d 159, 166 (Wis. 1988). Significantly, however, these decisions have generally rested on the courts’ recognition of a corollary duty to control a violent patient through involuntary commitment if necessary. See Lipari, 497 F. Supp. at 193 (holding that therapist’s duty includes “whatever precautions are reasonably necessary to protect potential victims of his patient,” including “duty to detain a patient” in hospital); Naidu, 539 A.2d at 1073 (holding that defendants had duty to warn “and a duty to control the actions of a mentally ill patient” and were negligent in discharging patient from hospital); Petersen, 671 P.2d at 237 (upholding judgment for plaintiff based on psychiatrist’s failure “to petition the court for a 90-day commitment, as he could have done ... to protect those who might foreseeably be endangered”); Schuster, 424 N.W.2d at 166 (rejecting defendants’ claim that they did not “have a duty to warn third parties or to institute proceedings for the detention or commitment of a dangerous individual for the protection of the patient or the public”). See generally M. Quattrocchi, Tarasaurus Rex: A Standard of Care That Could Not Adapt, 11 Psychol. Pub. Pol’y & L. 109, 113 (2005) (“Some courts have imposed a duty to third *377 parties in the absence of an identifiable victim. These cases emphasize . . . protection in the form of hospital confinement.”); R. Schopp, The Psychotherapist’s Duty to Protect the Public: The Appropriate Standard and the Foundation in Legal Theory and Empirical Premises, 70 Neb. L. Rev. 327 , 345 (1991) (noting that “the Schuster court interpreted warnings and civil commitment as comparable techniques for protecting the public from foreseeable harm”).
¶ 93. Thus, those courts that have broadened the therapist’s duty to all “foreseeable” victims without limitation have resolved the dilemma posed by the risk of over-commitment essentially by ignoring it; under these rulings, anyone injured by a mental health patient may argue that, in retrospect, the therapist was negligent in failing to detain the patient. In states like Vermont, however, where public policy militates against the recognition of a duty to control a patient through involuntary hospitalization — a policy reaffirmed by the majority today — extending the duty to the public at large is not a sound or practical option. See ante, ¶ 77 (rejecting imposition of duty to institutionalize mental health patient in order to avoid “an increase in unjustified commitments and abandonment of treatment-in-the-least-restrictive-environment” policy).
¶ 94. Most states, as noted, have pursued an approach between these two extremes. Through case law or legislation they have struck a balance among the competing concerns by recognizing a relatively narrow duty of care limited to situations where the therapist knows or should know that a patient poses a specific threat to an identified or reasonably identifiable third person. This standard, as one court has observed, “evinces a sound public policy against expanding the liability of health professionals to an indeterminate class of potential plaintiffs.” Eckhardt, 534 N.E.2d at 1345 . It reflects a considered policy judgment that the societal costs of breaching the therapeutic bond based on generalized threats of violence — all too commonplace in the therapeutic setting 24 — do not justify whatever uncertain benefits may flow from expanding the duty to unspecified third parties based on an inherently inexact risk assessment made all the more difficult *378 where the potential target is not identified. See, e.g., Thompson, 614 P.2d at 736 (observing that “it is fair to conclude that warnings given discreetly and to a limited number of persons would have a greater effect because they would alert [them] . . . of a specific threat pointed at them”).
¶ 95. This balancing of interests was cogently addressed by the Pennsylvania Supreme Court in considering “the conundrum a mental health care professional faces regarding the competing concerns of productive therapy, confidentiality and other aspects of the patient’s well being, as well as the interest in public safety.” Emerich v. Phila. Ctr. for Human Dev., Inc., 720 A.2d 1032, 1040 (Pa. 1998). In light of these concerns, the court concluded that the circumstances giving rise to a duty to third parties must necessarily be “limited,” requiring “the existence of a specific and immediate threat” which is “made against a specifically identified or readily identifiable victim.” Id. “Strong reasons,” the court concluded, compel the conclusion that the therapist’s duty “must have some limits.” Id. Many other courts have echoed these concerns in reaching similar conclusions. See, e.g., Fraser, 674 A.2d at 816 (adopting rule that therapist’s duty to protect third persons is limited to identifiable victims or class of identifiable victims based on “balance [of] the interests of those injured by psychiatric outpatients against the interests of the mental health profession in honoring the confidentiality of the patient-therapist relationship and in respecting the humanitarian and due process concerns that limit the involuntary hospitalization of the mentally ill” (citation omitted)); Eckhardt, 534 N.E.2d at 1345 (rejecting expansion of therapist’s duty beyond “cases involving specifically identifiable, potential victims as evidenced by specific threats” because “[h]uman behavior is simply too unpredictable and the field of psychotherapy presently too inexact,” and imposition of a broader duty “would be to place an unacceptably severe burden on those who provide mental health care to the people of this State, ultimately reducing the opportunities for needed care”). In addition, as noted, numerous states have codified similar, practical limits on a mental health care provider’s duty of care to third parties. See D. Mossman, Critique of Pure Risk Assessment or, Kant Meets Tarasoff, 75 U. Cin. L. Rev. 523 , 586 n.204 (2006) (observing that, “[t]o clarify clinicians’ responsibilities, many states have enacted laws that limit therapists’ potential liability if they take specified actions when a patient makes a serious threat *379 against an identifiable victim” (quotation omitted)); Nat’l Conference of State Legislatures, Mental Health Professionals’ Duty to Warn, www.ncsl.org/research/health/mental-health-professonals-duty-to-warn.aspx (2015) (collecting state statutes).
¶ 96. The point here is not that the Court has adopted a minority position without expressly acknowledging it. If that were the problem, it would be enough to simply articulate the competing viewpoint, and agree to disagree. However ill-advised the majority’s choice, it would at least be based on familiar ground. And while the decision to abandon a standard that so many states have found to be the proper balance between competing public-policy interests might be mistaken, it would at least have the virtue of transparency.
¶ 97. The problem here is altogether different, however, and far more serious. For the majority not only expands the scope of a therapist’s duty beyond the limits recognized by this Court in Peck, it creates an entirely new duty of care which plaintiffs here have labeled a duty to “train” and the majority sees fit to recast as “a duty to ‘inform’ that incorporates some elements of what plaintiffs describe as a distinct duty to ‘train.’ ” Ante, ¶ 53. Elsewhere, the majority variously describe this as: “a duty of care to provide reasonable information to the parents to enable them to recognize the dangers and fulfill the responsibilities envisioned for them in the treatment plan,” ante, ¶ 44; a duty that “contemplates more than simply advising the parents that their son posed a risk,” but also providing “reasonable information . . . to help keep their son safe,” ante, ¶ 53; and a duty to notify the caregiver of the risks “and of steps he or she can take to mitigate the risks,” ante, ¶ 82.
¶ 98. If mental health care providers, patients and their families, and the legal counsel who advise them remain uncertain as to the precise nature and scope of this new duty, they are to be forgiven. For the Court is making this up as it goes, with no input from the mental health profession on whether standards even exist for such a duty. This is not an easy task. But the real difficulty lies ahead, when this Court has moved on and the mental health care community must continue to grapple with the implications.
¶ 99. For make no mistake, this holding is extraordinary in its scope and implications. To recall, most duty-to-protect cases have divided along a fault-line between those limiting the duty to identified or reasonably identifiable targets of violence and those *380 that would include all “foreseeable” victims, the latter generally predicated on a duty to treat and, if necessary, confine a dangerous patient given the general impracticality of warning all remote, albeit foreseeable victims. The majority rejects the principle that a therapist’s duty extends only to reasonably identifiable targets of specific threats by the patient. It also rejects as a matter of policy any duty to control a mental health patient through involuntary commitment. Ante, ¶ 79.
¶ 100. Out of this seeming impasse the majority creates a new duty — a duty to warn not the patient or the patient’s targeted victims, but the patient’s parents or, more broadly, his or her “caretakers” so that they may control the patient and prevent injury to the public. This is worth a moment’s reflection. As a matter of policy, according to the majority, no liability may attach to E.R.’s mental health providers for their allegedly negligent failure to control E.R.’s conduct by providing for his involuntary commitment. Nevertheless, liability may attach to the same defendants for their allegedly negligent failure to enable E.R.’s parents to control his conduct by providing them with adequate warning and information to “mitigate the risks.” Ante, ¶ 82.
¶ 101. The imposition of a duty so novel and with such potentially broad consequences for mental health care providers, their patients, and the general public surely requires a more solid foundation than an allegation in a complaint. Recognizing that this case remains at the pleading stage, duty nevertheless constitutes an essential element of plaintiffs’ cause of action, and its existence is a question of law which this Court must decide in the first instance in light of all relevant policy concerns. See Endres v. Endres, 2008 VT 124, ¶ 11 , 185 Vt. 63 , 968 A.2d 336 (noting that “[d]uty ... is central to a negligence claim” and that “its existence is primarily a question of law” based on “those considerations of policy which lead the law to say that the plaintiff is entitled to protection” (quotation omitted)). Yet nothing in plaintiffs’ complaint even remotely identifies the basis for recognizing a so-called duty to “inform” a patient’s parents or “caretakers” to protect the public. Nothing in plaintiffs’ briefing below or before this Court identifies any medical treatises or other literature defining and describing the basic clinical standards, practices, and therapeutic goals underlying such a duty. Nothing in the briefing identifies any decisional law or authority elsewhere specifically recognizing and imposing such a duty.
*381 ¶ 102. We do, on the other hand, know from plaintiffs’ complaint that certain actions were taken by defendants after E.R.’s discharge. We know that the Brattleboro Retreat made an aftercare treatment plan for E.R. and reviewed it with E.R. and his parents, and that the plan “involved E.R. being seen on a regular basis” at NKHS. We know that the medical professionals at the Retreat prescribed medications for E.R. to take on a daily basis. We know that E.R. met with a “treatment team” at NKHS, and that a “cognitive remediation therapy” plan was put in place and signed by E.R. We know that E.R told his mother in mid-December 2010 that he had ceased taking his medications, and that she reported this to NKHS. And we know that, on the day of the assault, E.R.’s father had taken E.R. with him to oversee work being done at an apartment building owned by E.R.’s grandfather.
¶ 103. These facts themselves, hardly unique, highlight the most significant deficiency in the majority’s newfound duty. Even assuming that plaintiffs could establish through expert evidence some professional standards for the duty owed to a patient’s caretaker, the imposition of such a duty demands consideration of the policy implications underlying it — its practical benefits against its societal costs. See Langle v. Kurkul, 146 Vt. 513, 519 , 510 A.2d 1301, 1305 (1986) (noting that existence of duty is primarily question of law dependent on variety of policy concerns, including “the closeness of the connection between the defendant’s conduct and the injury suffered,” the “burden to the defendant,” and the “consequences to the community” (quotation omitted)).
¶ 104. The facts alleged by plaintiffs show that, even with the practical steps undertaken by defendants and E.R.’s parents to facilitate his functioning safely in a less restrictive environment than a closed hospital ward — providing an aftercare plan and reviewing it with E.R.’s parents, establishing an outpatient treatment team, prescribing him daily medication, endeavoring to monitor his activity by taking him to job sites, and reporting that he had stopped his medication — they could not prevent him from perpetrating a spontaneous act of violence.
¶ 105. The majority’s speculation that defendants might have provided some additional “information” to E.R.’s parents to prevent the assault simply misses the point. To impose such a duty on health care providers undermines the fundamental policy underlying our mental health care system, a policy designed to *382 maximize a patient’s freedom and dignity by providing treatment in the least restrictive environment available. It is the same policy resoundingly reaffirmed by the majority today in refusing to impose a tort duty on health care providers to institutionalize a patient. See ante, ¶ 77 (rejecting duty to institutionalize mental health patient to avoid “an increase in unjustified commitments and abandonment of treatment-in-the-least-restrictive-environment” policy).
¶ 106. Uncertainty counsels caution, for courts and clinicians alike. Any responsible mental health care provider uncertain as to how, if at all, to satisfy this new, amorphous duty to train or assist a patient’s “caretaker” sufficiently to prevent future harm might understandably decide to err on the side of a more — rather than a less — restrictive treatment setting rather than risk a lawsuit by the random victim of an outpatient assault. Moreover, considering the many adult patients living with someone who could be characterized as a “caretaker” — be it the patient’s parents, spouse, domestic partner, or friend — the consequences of such decisions could be far reaching. Balanced against the dubious odds of actually predicting, much less preventing, random acts of violence by a patient absent any specific threat or identifiable victim, the risk becomes prohibitive.
¶ 107. The expected response to these concerns is that they are merely “speculative” while we know — in contrast — that mental health providers routinely make predictions of dangerousness in deciding to commit a patient and routinely apply Tarasoff when deciding whether a patient poses a threat. Thus, it is easy to posit that the concern for overcommitment is exaggerated or unfounded, that no responsible mental health care provider would involuntarily hospitalize a nondangerous patient to avoid a lawsuit, much less release a dangerous one despite the risk to the public.
¶ 108. The flaw in this response is the assumption that there are “yes” or “no” answers to the mental health clinician’s decisions. The relevant medical and legal literature, however, belies this assumption. There are, in fact, no answers, but only imperfect assessments of differential levels of risk, and there are no clear standards defining the level of risk sufficient to trigger protective measures. See, e.g., Mossman, supra, 75 U. Cin. L. Rev., at 567, 577 (observing that most recent medical studies show that “a therapist’s predictive knowledge about future violence is really an ability to make risk estimates,” while “there is and can be no *383 rationally established, broadly accepted criterion for what probability of risk constitutes the level of ‘serious danger’ that should trigger a protective response”); Herbert, supra, 30 J. Am. Acad. of Psychiatry & Law, at 422 (noting the “residuum of uncertainty” in assessing whether “a patient really means particular words as a threat”). Tarasoff has worked, according to surveys and studies, because most states employ reasonably clear, narrow, and understandable standards that require a serious threat to a reasonably identifiable target. See, e.g., Rosenhan, supra, 24 Pac. L.J., at 1203, 1208, 1217 (findings from broad survey of psychotherapists showed that, while very small percentage rated their ability to assess dangerousness “very accurately,” most understood duty to protect was predicated on identification of specific victim and believed that duty was consistent with ethical obligations); Mossman, supra, 75 U. Cin. L. Rev., at 603 (noting that clinicians have sought and been well served by “statutory boundaries on the duty to protect, boundaries that tell them when the duty arises (usually, following explicit threats toward specific targets) and that define specific ways of discharging the duty”); M. Soulier, et al., Status of the Psychiatric Duty to Protect, Circa 2006, 38 J. Am. Acad. of Psychiatry & Law 457, 471-72 (2010) (concluding from surveys of psychotherapists and review of legal evolution of Tarasoff duty that “statutes appear to promote a useful social policy, limiting the duty to protect to cases in which victims are identified or reasonably identifiable” and as such pose little threat to clinician’s ability to practice). The broad duty created by the majority, in contrast, contains none of the limits that form a natural and necessary counterbalance to the risks of defensive practice and overcommitment in the mental health context.
¶ 109. To dismiss the concerns of the mental health care profession in this case as speculative or even self-serving, moreover, is presumptuous. It is all too easy to assign new duties to a profession we know little about, and have no responsibility to implement. Judicial restraint in creating duties for other professions is not an end in itself; it is the end-result of recognizing our own limitations. It is wisdom grounded in humility.
¶ 110. The majority ultimately attempts to minimize its decision’s impact by noting the “significant obstacles” in the way of any successful lawsuit, including the plaintiffs’ need to prove the “necessary facts,” a failure to disclose the necessary information, as well as “causation — that any failure to inform . . . was more *384 likely than not a but-for cause of their injuries.” Ante, ¶ 83. This may reassure the majority, but it is cold comfort to the mental health care providers and their colleagues and families compelled to endure the personal and professional disruptions, stress, and financial burdens of protracted lawsuits predicated on this amorphous new duty, regardless of their ultimate success.
¶ 111. Finally, I would note that the majority’s alternative basis for imposing a duty of care predicated on its conclusion that “E.R.’s parents fell within the ‘zone of danger’ from E.R.’s conduct” is equally flawed and unpersuasive. Ante, ¶ 47. The zone-of-danger doctrine, as noted, simply extends the therapist’s duty to persons within a finite class of reasonably identifiable potential targets. Thus, in the case cited by the majority, Hammam v. County of Maricopa, the record showed that the patient had “expressed jealousy of his stepfather” to the therapist; that the patient’s parents had expressed concern to the therapist for their safety and begged the therapist to admit the patient to the hospital; that the therapist failed to do so; and that the patient subsequently attacked his stepfather with an electric drill. 775 P.2d at 1123-24. Based on these facts, the court reasonably concluded that, despite the absence of a specific verbalized threat against the parents, they “were readily identifiable persons who might suffer harm.” Id. at 1128.
¶ 112. Despite the majority’s statement that it “find[s] Hammam persuasive and follow[s] its reasoning,” ante, ¶ 49, nothing on the limited factual record here brings this case within the “zone of danger” doctrine articulated in Hammam and elsewhere. First, the complaint did not allege that E.R. had threatened either his parents or a class of persons that might reasonably be construed to include his parents. 25 Nor did plaintiffs claim, as the majority argues, that E.R.’s earlier aggression toward a member of the staff at the Retreat somehow brought E.R.’s parents into the zone of danger applicable to all “caretakers.” To suggest that a threat against a nurse, therapist, physician or other mental health care provider somehow represents a threat against an identifiable class of all family members and friends who help with the patient’s outpatient care would stretch the “zone of danger” doctrine beyond recognition.
*385 ¶ 113. Second, and more significantly, the doctrine was designed to protect a slightly expanded class of reasonably identifiable potential victims, and E.R.’s parents were not the victims here. As noted, plaintiffs did not allege any threats — explicit, implicit, or otherwise — against his parents. Nor is there any factual basis to support a conclusion that the actual victim, Mr. Kuligoski, was within an identified or identifiable class of potential victims. The “zone of danger” argument thus fails entirely.
¶ 114. This Court has repeatedly cautioned against placing “our imprimatur” upon a new legal duty “without first determining whether there is a compelling public policy reason for the change.” Langle, 146 Vt. at 520 , 510 A.2d at 1306 ; accord Goodby v. Vetpharm, Inc., 2009 VT 52, ¶ 11 , 186 Vt. 63 , 974 A.2d 1269 ; Knight v. Rower, 170 Vt. 96, 107 , 742 A.2d 1237, 1245 (1999); Smith v. Day, 148 Vt. 595, 599 , 538 A.2d 157, 158 (1987). The majority identifies no compelling public policies to warrant the extraordinary duty it imposes on mental health care providers by today’s ruling. On the contrary, settled public policy governing our treatment of the mentally ill demands precisely the opposite result. I therefore respectfully dissent.
¶ 115. I am authorized to state that Justice Skoglund joins this dissent.
R. Traynor, Transatlantic Reflections on Leeways and Limits of Appellate Courts, 1980 Utah L. Rev. 255 , reprinted in The Traynor Reader 200 (1987).
Traynor, supra, at 200.
In Thompson v. County of Alameda, 614 P.2d 728, 734 (Cal. 1980), the California Supreme Court clarified Tarasoff by explaining that a therapist’s duty to protect arises only when the patient’s intended victim is “readily identifiable.” “[NJonspe-cific threats of harm directed at nonspecific victims” do not trigger the duty of care. Id. at 735 .
As discussed more fully below, a few courts have expanded the duty slightly to include persons within a “zone of danger” who were sufficiently targeted by the patient even if not specifically threatened. See, e.g., Jablonski v. United States, 712 F.2d 391, 398 (9th Cir. 1983) (applying California law and Tarasoff to hold that, although defendant’s patient had made no express threat against his domestic partner, Melinda Kimball, she was within scope of duty where patient’s “previous history indicated that he would likely direct his violence against Kimball,” his psychological profile “indicated that his violence was likely to be directed against women very close to him,” and he had threatened Kimball’s mother), overruled on other grounds by In re McLinn, 739 F.2d 1395 (9th Cir. 1984) (en banc); Hamman v. Cty. of Maricopa, 775 P.2d 1122, 1127-28 (Ariz. 1989) (holding that “Tarasoff envisioned a broader scope” of duty than circumstance where patient “verbalized [a] specific threat,” and could include patient’s family where his threats placed them “within the zone of danger, that is, subject to probable risk of the patient’s violent conduct”); see also Fraser, 674 A.2d at 816 (noting that most courts have *375 extended therapist’s duty of care only to “victims who were . . . either specifically identifiable or within a class of foreseeable victims”).
See, e.g., D. Rosenhan, et al., Warning Third Parties: The Ripple Effects of Tarasoff, 24 Pac. L.J. 1165 , 1222 (1993) (concluding, based on survey of mental health providers, that in accordance with Tarasoff “psychotherapists continue to warn patients that certain conversation is not confidential,” and “in accord with expectation, many of these patients simply abandon treatment,” posing additional risks to the public).
The clinical difficulties in (1) assessing the risk of violence posed by a patient and (2) determining whether that risk is sufficient to warrant protective actions, recognized in Tarasoff, have not appreciably lessened in the decades since. See, e.g., D. Mossman, Critique of Pure Risk Assessment or, Kant Meets Tarasoff, 75 U. Cin. L. Rev. 523 , 601-02 (2006) (explaining that clinicians do not “predict dangerousness” but simply identify different “levels of risk,” and that more significantly few empirical studies reveal “what level of risk is sufficient to justify . . . action”); P. Herbert, The Duty to Warn; A Reconsideration and Critique, 30 J. Am. Acad. of Psychiatry & Law 417, 421 (2002) (observing that, “despite advances in risk assessment,” such assessments fall “substantially short of exact science” and involve at best “approximations of the degree of risk”).
This concern was cogently summarized by the court in Sherrill v. Wilson, 653 S.W.2d 661, 664 (Mo. 1983):
The plaintiff could undoubtedly find qualified psychiatrists who would testify that the treating physicians exercised negligent judgment, especially when they are fortified by hindsight. The effect would be fairly predictable. The treating physicians would indulge every presumption in favor of further restraint, out of fear of being sued. Such a climate is not in the public interest.
See Herbert, supra, 30 J. Am. Acad. of Psychiatry & Law, at 422 (explaining that “mental health workers must grapple with threats of suicide or of violence against others regularly as an integral part of their work,” and that such threats “are daily grist”).
At the motion hearing, plaintiffs’ counsel readily conceded that ‘There was no identifiable victim” in this case.

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/cases/9836153. Public record. Not legal advice.
