Currie v. United States

644 F. Supp. 1074, 55 U.S.L.W. 2211, 1986 U.S. Dist. LEXIS 19491

Opinion

MEMORANDUM OPINION

GORDON, Senior District Judge:

This case comes before the court on defendant’s motion to dismiss, or in the alternative for summary judgment. While the court finds that a claim for relief is stated by plaintiff’s claim, it nevertheless finds the facts inadequate as a matter of law to support the limited claim accepted by the court. Therefore, the summary judgment motion of the defendant will be granted.

FACTS

At 1:30 p.m. on August 30,1982, Leonard Avery, a Vietnam veteran suffering from Post-traumatic Stress Disorder, entered a building containing the IBM medical department at Research Triangle Park, North Carolina. Dressed in army fatigues and carrying a .45 caliber semi-automatic rifle and homemade bombs, Avery shot and killed Ralph Augustus Glenn, Jr., and wounded several other persons. Avery subsequently was convicted of first degree murder.

Avery, an IBM employee, had been treated on an outpatient basis at the Mental Hygiene Clinic of the Veterans Administration Hospital in Durham, North Carolina since April 1, 1981. He had originally consulted the VA staff due to “rage attacks” he had been increasingly experiencing, including one incident in which he left work because he feared “losfing] control” due to anger at a fellow employee and another incident when he became angry at a neighbor and started shooting a gun.

Avery was given a psychological profile test in May of 1981 and in July was diagnosed as suffering from Post-traumatic Stress Disorder. VA doctors referred him to group therapy sessions and prescribed various types of anti-psychotic medication such as Mellaril and Haldol.

On October 28, 1981, Avery requested an emergency meeting with Dr. Buck, his group leader at the VA Hospital, after becoming upset over a child support dispute with his former wife. Dr. Buck’s examination revealed that Avery had operated at a “marginal level” for several months, but that his emotional condition had deteriorated such that he presented a “significant homicidal risk.” Avery admitted to carrying a gun “with thoughts of hurting anyone who would take away his property.” Avery voluntarily admitted himself to the VA Hospital and, after telephone conversations with his attorney and ex-wife calmed him down, was released at his request the next day.

Avery also had problems at IBM. The noises and actions of co-workers often upset him. In November 1981, Avery stayed out of work for several weeks without notifying his supervisor. IBM allowed him time off with pay when absences were medically excused, and consequently Avery was not discharged when Dr. Buck’s subsequent examination led him to write IBM that Avery’s absences had been justified. During 1982, Avery continued to experience problems at work, such as when he left work for two days after becoming angry and upset at the presence of Oriental visitors.

During this time, Avery continued to attend group sessions, although somewhat sporadically. In July 1982, Avery’s demeanor at the sessions and reports of Avery’s emotional problems at work prompted Dr. Buck to begin writing one-week medical excuses each week at the group therapy session. At the last session Avery attended, July 26, 1982, he was is sued a work excuse valid through August 2.

Avery failed to attend group sessions or work after August 2,1982. On August 13, IBM doctors consulted Dr. Buck and learned that Avery, who had telephoned IBM and claimed medical excuses for his absences, had actually been excused only through August 2. Nevertheless, IBM apparently decided to excuse Avery if he attended the August 16, 1982 group session.

When Avery failed to attend that session, he was called to IBM on August 18 to explain his absences. Avery responded by threatening to blow up IBM’s medical facility and personnel. IBM notified Dr. Colvard at the VA Mental Hygiene Clinic of these threats. Dr. Colvard advised IBM that Avery did have the potential for such acts of violence and that IBM should take Avery’s threats seriously and contact the appropriate law enforcement authorities. Soon thereafter, Dr. Buck telephoned Avery. Avery agreed to voluntarily commit himself, but he never showed up at the facility. Upon learning of this fact, IBM called Avery on August 19 or 20 and told him he was fired.

On August 20, Avery called Dr. Buck and the following conversation ensued:

Avery: I’ve been fired.

Dr. Buck: Yes, I know, Dr. Connor [at IBM] told me.

Avery: I want to see you. I want you to write a letter saying it was o.k. to be out of work.

Dr. Buck: I don’t see how I can write you a letter. IBM knows that you haven’t been coming for your appointments when you were saying that you were.

Avery: I tried to call you.

Dr. Buck: There were no messages left in my box until this last Wednesday.

Avery: It's too late for me to start over. There’s nothing to be done. You’ll read about it in the papers.

Dr. Buck interpreted Avery’s last statement as a threat and immediately called IBM to warn of the threat. He also discussed with two other VA psychotherapists the possibility of involuntarily committing Avery and all agreed that Avery should not be committed. Because Avery appeared to be articulate and in touch with reality during his phone conversation with Dr. Buck, the psychotherapists concluded that Avery's problems arose from his anger at IBM at having discovered his abuse of medical leave rather than from mental illness. Because they did not believe Avery to be "mentally ill,” they did not believe involuntary commitment was feasible under North Carolina law, which requires that patients be both dangerous and mentally ill.

When Avery failed to attend the next group session on August 23, the VA, at IBM’s request, sent Avery a postcard asking Avery to make an appointment at the VA Clinic. That same day, at a weekly staff meeting for the Mental Hygiene Clinic, eight psychiatrists, all with faculty positions at Duke University Medical School and seven with board certification in psychiatry, discussed Avery’s case and unanimously concluded that Avery could not be committed under North Carolina law.

On August 26, Avery called the VA Hospital in response to the postcard sent to him. When Dr. Buck returned his call later in the day, the following conversation transpired:

Dr. Buck: Mr. Avery? It’s Dr. Buck. I called to see how you are doing.

Avery: I’m doing all right.

Dr. Buck: Will you be coming to the group [session]?

Avery: I’m not going back to that group. You sons of bitches never did anything to help me. (Silence) I can find you.

Dr. Buck: I’m sorry. What did you say? (Silence) Mr. Avery, are you making a threat? (Silence)

Avery: I’d like to get all you guys in the same place at the same time. I’d blow your asses away.

Dr. Buck: Mr. Avery, I hope you’re not going to do anything to hurt someone or get yourself in trouble with the law.

Avery: I don’t have time to talk to you.

Dr. Buck: I hope things go better for you. Goodbye.

Dr. Buck immediately notified IBM of this conversation. He also again informally discussed with resident psychotherapists the possibility of involuntarily committing Avery, but no action was taken.

On August 30, another weekly staff meeting was held and the psychiatrists again agreed that Avery could not be committed. The staff members also discussed any obligations they might have under Tarasoff v. Regents of the University of California, 17 Cal.3d 425, 131 Cal.Rptr. 14, 551 P.2d 334 (1976) and concluded that, because the United States Attorney, the VA District Counsel and police, the Federal Bureau of Investigation, and the Durham County Police Department, as well as IBM, had been notified and given the necessary data on Avery and his threats, those obligations had been fulfilled.

A few hours later, Ralph Augustus Glenn, Jr. was dead.

Following an unsuccessful administrative claim, plaintiff brought this action alleging that the VA doctors were negligent in failing to petition to involuntarily commit Avery. Plaintiff does not contend that the VA doctors failed to fulfil any duty to warn that might be imposed on them by law.

DISCUSSION

Plaintiff brings .his claim under the Federal Tort Claims Act, 28 U.S.C. § 1346(b) (1982), which provides that the United States may be civilly liable to the same extent as private persons for personal injury caused by the negligent acts of its employees acting within the scope of their employment. This limited waiver of sovereign immunity subjects the United States to a state’s tort law. Section 1346(b) also mandates that the proper choice of law is that of the state where the alleged tort occurred. Here, there is no question that North Carolina law applies. Since North Carolina has not directly considered the issue in this case, this court will look to other jurisdictions for persuasive precedent before examining those decisions in the context of North Carolina law.

A. The Tarasoff Decision and North Carolina Law

The seminal case in this area, and the case from which all courts examining a duty to commit have begun' their examination, is Tarasoff v. Regents of the University of California, 17 Cal.3d 425, 131 Cal.Rptr. 14, 551 P.2d 334 (1976). In Tarasoff, a therapist was found liable for failing to warn an unnamed but readily identifiable victim that the therapist’s patient had threatened her.

The Tarasoff case was significant in three important respects: (1) it found that a psychotherapist and patient had a “special relationship” that could impose on the psychotherapist an affirmative duty to exercise reasonable care, (2) it found that policy considerations justified imposing a duty to warn as a component of reasonable care when a psychotherapist knew or should have known that a patient was dangerous, and (3) it found that therapists owed this duty at least to readily identifiable third persons whom the patient had threatened.

It is believed that North Carolina would adopt Tarasoff s affirmative duty on therapists to warn foreseeable third parties. A North Carolina Court of Appeals opinion, In re Farrow, 41 N.C.App. 680, 255 S.E.2d 777 (1979), has cited Tarasoff with seeming approval when discussing a psychotherapist’s “duty to disclose when necessary to protect his patient or the public from imminent danger.” Farrow, 255 S.E.2d at 781 (emphasis added). Similarly, in Pang-burn v. Saad, 73 N.C.App. 336, 326 S.E.2d 365 (1985), the court quoted the following from a Georgia case which had followed Tarasoff:

“[Wjhere the course of treatment of a mental patient involves an exercise of ‘control’ over [the patient] by a physician who knows or should know that the patient is likely to cause bodily harm to others, an independent duty arises from that relationship and falls upon the phy sitian to exercise that control with such reasonable care as to prevent harm to others at the hand of the patient.”

Pangbum, 326 S.E.2d at 367 (quoting Bradley Center, Inc. v. Wessner, 250 Ga. 199, 201, 296 S.E.2d 693, 695-96 (1982)). 1 Finally, the court notes that the vast majority of courts that have considered the issue have accepted the Tarasoff analysis.

B. Distinctions from Tarasoff

Despite the virtually universal judicial approval of the Tarasoff decision itself, there have been numerous distinctions drawn from its holding. Defendant here claims that even if North Carolina were to adopt the Tarasoff rule, it would similarly distinguish and restrict its ruling. Defendant claims that the instant case is distinguishable because (1) plaintiff was not a “readily identifiable” victim here, as in Tarasoff, and (2) plaintiff here alleges that the therapist had a duty to seek his patient’s involuntary commitment rather than merely a duty to warn those whom his patient had threatened, as in Tarasoff.

1. “Readily Identifiable” Victim

The Tarasoff decision was subsequently limited by the California Supreme Court in Thompson v. County of Alameda, 27 Cal.3d 741, 167 Cal.Rptr. 70, 614 P.2d 728 (1980). In Thompson, a juvenile offender who was confined in a county institution made general threats that he would kill a child in his neighborhood; while on temporary leave he made good on his macabre threat. In a subsequent Tarasoff- like action for failure to warn of the patient’s threats, the California Supreme Court upheld a dismissal of the claim. The court reasoned that unlike Tarasoff, in which the victim was unnamed but still “readily identifiable,” the victim in Thompson was not clearly known but instead one of several children in the mental patient’s neighborhood. The court reasoned that the lack of a specific threat made the threat less credible and also made warnings more difficult and less effective as a practical matter.

In the subject case, defendant alleges that Avery’s threats were made only to IBM, and not specifically to the ultimate victim. Defendant notes that the IBM class of which Mr. Currie was a part arguably is no more specific than the neighborhood of which the victim in Thompson was a member.

This court, however, does not believe that the “readily identifiable victims” rule is sufficiently encompassing to conform with basic tort law or with North Carolina law in the area of assessing a duty to involuntarily commit patients. For this reason, the Thompson rule will not be accepted in this case.

We agree with the concurring opinion in Vu v. Singer Co., 706 F.2d 1027 (9th Cir.), cert. denied, 464 U.S. 938,104 S.Ct. 350, 78 L.Ed.2d 315 (1983), that limiting a defendant’s duty to readily identifiable victims represents an overly restrictive view of foreseeability. Id. at 1031 (Rothstein, J., concurring). See Lipari, 497 F.Supp. 185, 194 (D.Neb.1980); Petersen v. State, 100 Wash.2d 421, 671 P.2d 230, 237 (1983) (en banc). See also Note, 19 New Eng.L.Rev. 597 (1984) (advocating basic foreseeability standard); Note, 58 St. John’s L. Rev. 492, 500 & n. 32 (1984) (identifiable victim limitation “stands in contrast with the broader test of foreseeability that normally applies in negligence actions”; Palsgraf cited); Note, 76 Nw. U.L.Rev. 331 (1981) {Thompson is “novel approach” to duty analysis). Clearly there are victims not “readily identifiable” but nevertheless “foreseeable.” As noted by the dissenting opinion in Thompson, it is contrary to basic tort principles “to confine that duty [owed by a doctor treating a patient whose medication made him a dangerous driver] to motorists or pedestrians whom the doctor could identify in advance.” Thompson, 167 Cal.Rptr. at 82, 614 P.2d at 740 (Tobriner, J., dissenting).

This discussion is not necessarily saying that Thompson was wrongly decided. It is instead saying that Thompson’s limitation should not be viewed as a foreseeability restriction on duty, but rather as a policy restriction on duty. Just as the California Supreme Court had decided in Tarasoff that the benefits of warning a readily identifiable victim outweighed the risks of unnecessary warnings, it decided in Thompson that the benefits of warning nonspecific victims about nonspecific threats did not outweigh these risks.

The importance of viewing the rule in Thompson as a policy restriction rather than as a foreseeability restriction becomes particularly evident when the Thompson rule is evaluated in the context of a duty to involuntarily commit case. This court believes that the policy considerations in a duty to commit case differ significantly from a duty to warn case and concludes that North Carolina would not limit a therapist’s “duty to commit” to patients who have “readily identifiable” victims.

One cannot read Thompson without being struck by the emphasis the court placed on the practical difficulties which the lack of a “readily identifiable” victim placed on a duty to warn. Without knowing specifically who the victim would be, the therapist could not know who to warn. The court discussed the possibility of a general warning, but noted that the effectiveness of such a warning would be greatly reduced because people would pay less attention to them. 167 Cal.Rptr. at 77-80, 614 P.2d at 735-38. In sum, the court found that, unlike in Tarasoff, Thompson’s reduced benefits of general warnings would not outweigh the problems such as breach of patient confidentiality, now worsened by broader dissemination of his threats.

This conclusion cannot be reached in a “duty to commit” balancing test. Unlike a duty to warn case, in which the therapist needs to know the identity of the victim in order to adequately act, the therapist in a duty to commit case need only know that the patient is dangerous generally in order to adequately commit him. As a practical matter, the victim’s identity is irrelevant to whether the doctor can adequately act — by committing the patient, the therapist is able to protect all possible victims.

The court does not believe that it is wise to limit any duty to commit according to the victim. 2 Arguably, the patient who will kill wildly (rather than specifically identifiable victims) is the one most in need of confinement. In negligent release cases, a defendant’s duty generally has not been limited to readily identifiable victims, see Pangburn, and the court believes a similar rule is appropriate here. Citizens outside of the “readily identifiable” sphere but still within the “foreseeable zone of danger” are potential victims a therapist should consider if he has a duty to them and a means of adequately protecting them. See Stone, 90 Harv.L.Rev. 358, 375-76 (1976).

In the instant case, Avery made clear threats against the IBM medical facilities. The VA Hospital Doctors told IBM to take these threats seriously. While the ultimate victim arguably was not “readily identifiable,” he was clearly within the “foreseeable zone of danger.”

Thus, this court finds that the victim, Mr. Glenn, fits within the “foreseeable victims” rule. It also finds that the “readily identifiable” victims rule of Thompson was a policy limitation on a therapist’s duty to warn based on practical difficulties not present in a duty to commit case. Thus, the court will not dismiss plaintiff’s claim based on the status of the victim in this case.

2. Duty to Commit v. Duty to Warn

The court now comes to the most difficult part of its decision: whether an affirmative duty to commit should exist at all. Plaintiff would have the court find that, once a special relationship is found, the therapist should exercise “reasonable care under the circumstances,” which plaintiff contends is a question for the jury. Plaintiff cites in support of this position Restatement (Second) of Torts § 315(a), which states that once such a relationship is found, a party has an affirmative duty “to control the third person’s conduct.”

Plaintiff is correct that Tarasoff s finding of a special relationship between therapist and patient allows an affirmative duty to be placed on a party “to control” another. But the court believes that plaintiff’s proposed standard of “reasonable care under the circumstances” begs the question. What is “reasonable” depends on a court’s review of the policy considerations present in that case. For example, Tarasoff found a special relationship between therapist and patient and then stated that the relationship imposed on therapists a duty to take reasonable steps, which would vary according to the circumstances. It did not say that because of the relationship, courts should simply give the case to the jury and ask “what should a psychotherapist reasonably have done?” Instead, the Tarasoff court went on to examine whether policy considerations justified classifying a duty to warn as a type of reasonable care that society could expect a psychotherapist to perform. 3 These statements are consistent with the idea that “duty” is a question for the court — the court defines the parameters of reasonable conduct given the societal considerations, and the jury decides whether the facts of the case fit within those parameters.

Defining the parameters of “duty” necessarily involves a balancing of policy considerations. Prosser & Keeton on Torts § 54, at 358 (5th ed. 1984). In focusing on policy considerations, Tarasoff mentioned two risks of imposing a duty to warn. Warnings breach the therapist-patient privilege, and thus threaten (1) effective treatment, because patients might be less open if they knew their therapists might notify authorities or the victim, and (2) the patient’s right to privacy. These risks are exacerbated when the issue is a duty to commit: (1) If patients are inhibited by the possibility of warnings every time they mention a threat, they likely will be more inhibited by the realization that a therapist might seek to involuntarily commit them for fear of a lawsuit, and (2) while one might argue that disclosure of the threat through warnings invades a patient’s privacy more than commitment, the Tarasoff court expressly found otherwise. 131 Cal.Rptr. at 27 n. 14, 551 P.2d at 347 n. 14. In addition, an even more important factor is present in a duty to commit case: (3) loss of liberty, which raises societal harms of constitutional proportions. 4

The number of threats made by mental patients exceeds by a vast margin the number that represent serious threats. Tarasoff, 131 Cal.Rptr. at 40, n. 5, 551 P.2d at 360 n. 5 (Clark, J., dissenting). Accord Miller & Fiddleman, 60 N.C.L.Rev. 985, 990 n. 22 (1982). Even a relatively minor change in the method by which these threats are evaluated could significantly increase confinement and raise serious constitutional concerns. See O’Connor v. Donaldson, 422 U.S. 563, 575, 95 S.Ct. 2486, 2493-94, 45 L.Ed.2d 396 (1975) (unconstitutional to confine nondangerous mentally ill who can live safely in freedom). North Carolina’s stated policy, in fact, is to “favor a less restrictive mode of treatment than involuntary commitment whenever appropriate.” Farrow, 255 S.E.2d at 782.

The vast majority of courts have dismissed claims brought against therapists for failure to commit, based either on the fact that the victim was not “readily identifiable,” discussed supra, or on these policy grounds. Only one case clearly authorizes a claim for a duty to commit patients, Lipari v. Sears, Roebuck & Co., 497 F.Supp. 185 (D.Neb.1980), while two others can be read as arguably supporting such a theory. Durflinger v. Artiles, 234 Kan. 484, 673 P.2d 86 (1983), quoted in 727 F.2d 888 (10th Cir.1984); Petersen v. State, 100 Wash.2d 421, 671 P.2d 230 (1983) (en banc).

This court, however, agrees with Lipari that it would be improper to rule that a psychotherapist would never have a duty to institute commitment proceedings against a patient. To rule otherwise would allow a psychotherapist to act in “careless disregard” of members of the public known to be endangered by his patient. Tarasoff, 131 Cal.Rptr. at 22, 551 P.2d at 342. The court believes that a psychotherapist, perhaps the only one with knowledge of the danger posed by his patient, may have a duty to protect society by taking the only practical action he can — which may include seeking the involuntary commitment of a patient whom he knows is dangerous. 5

Policy considerations support the imposition of such a rule. Liability for a patient’s actions is generally recognized for negligent release from a hospital. Pangburn; Restatement (Second) of Torts § 319 (illustration 2). Much has been made of the distinction between negligent release cases and failure to commit cases. This distinction seems to this court, however, to be largely due to the common law presumption against imposing affirmative duties. Admittedly, liability for negligent release is clear, because no affirmative act is required. But liability for failure to commit is also possible once the “special relationship” exception has been found — the only question is whether policy considerations support the imposition of such a duty. Cf. Tarasoff, 131 Cal.Rptr. at 27 n. 13, 551 P.2d at 347 n. 13. Whether in a negligent release or involuntary commitment case, the basic societal goal is the same: dangerous mentally ill patients should be restrained, nondangerous capable of existing in society ones should be free. See N.C. GemStat. § 122C-201 (Cum.Supp.1985).

It is true that anyone may file a petition to involuntarily commit a patient; it may be argued that it is unfair to impose this duty on psychotherapists that is not imposed on the general public. But this argument is what is embodied in the concept of a special relationship between therapist and patient. See, e.g., Abernathy, 773 F.2d at 189 (fact that none of defendant’s agents were psychotherapists contributed to holding of no liability). Psychotherapists are more likely than the general population to know of their patients’ dangerous propensities, and the North Carolina courts have noted on several occasions the important role that psychotherapists play in the commitment procedure. E.g., Farrow, 255 S.E.2d at 780 (“Legislature intended for [patient’s] physician to play a key role almost from the inception of the proceedings”). North Carolina has attempted to encourage psychotherapist participation: for example, unlike other citizens, physicians who petition for involuntary commitment need not appear before the clerk or magistrate. Id. Ten percent of the involuntary commitments in North Carolina are initiated by therapists, Miller & Fiddleman, 60 N.C.L.Rev. at 994 n. 37, and North Carolina courts have stated that “[i]n a real sense the [involuntary commitment] proceeding is an important step in [a patient’s] medical and psychiatric treatment.” Farrow, 255 S.E.2d at 780. “Clearly the Legislature contemplated that a physician who has knowledge of a mentally ill or inebriate person who is ... dangerous to himself or others should initiate the proceedings for such person’s commitment by executing the oath setting forth his knowledge of the facts.” Id. (emphasis added). In sum, this court does not believe that North Carolina courts would allow its psychotherapists to make decisions whether to petition for a dangerous patient’s involuntary commitment “in careless disregard of [others’] life and safety.” Tarasoff, 131 Cal.Rptr. at 22, 551 P.2d at 342.

This court also believes that this rule will not have an unduly adverse backlash, leading to any significant risk of overcommitment by psychotherapists seeking to avoid potential liability. The harm in overcommitment is clearly a grave one, and one which this court cannot overlook. Nevertheless, the court believes the risk of this harm is mitigated due to several factors. First, defendant has shown no empirical evidence of overcommitment resulting from Lipari or the negligent release cases. See Goodman, 3 Behav.Sci. & L. 195, 219 (1985) (“There has been no creditable evidence either that the practice of psychotherapy has suffered or that violence within our society has increased because of the imposition of the duty to protect others upon the mental health professions.”). Second, a psychotherapist’s decision to petition for involuntary commitment does not settle the issue — that decision will be reviewed by a magistrate and two other psychiatric personnel, with full constitutional safeguards afforded. Third, for a therapist to be liable, a victim still must show the normal requirements for liability. Many of the attacks on reviewing therapists for liability often seem based as much in a lack of confidence in our jury system at large as in any substantive argument. Finally, the policy arguments against imposing liability on therapists must be weighed against the policy arguments in favor of our tort system generally. White v. United States, 780 F.2d 97, 103 (D.C.Cir.1986). These include compensating innocent plaintiffs vis-a-vis more culpable defendants and deterring future harmful conduct.

Thus, the court believes that therapists have some duty not to let known dangerous mental patients whom they treat run around in public. Cf. Tarasoff (“The containment of such risks lies in the public interest”); Pangburn, 326 S.E.2d at 371 (public employees “serve the needs of society as a whole as well as the needs of individual persons”).

At the same time, however, the court cannot help but recognize the risk of stepping off the precipice here. Even assuming that juries will always decide correctly, the perception of therapists that they will be second-guessed could significantly influence their decisions to commit. Cf. Azzolino v. Dingfelder, 315 N.C. 103, 337 S.E.2d 528, 535 (1985) (potential that wrongful life liability would “place increased pressure upon physicians to take the ‘safe’ course” cited as justification for refusing to recognize such a claim). Even absent empirical data, the court would have to close its eyes to ignore the potential harm to the recognized societal goal of minimum necessary confinement for the mentally impaired. 6 And while patients are given constitutional safeguards at the commitment hearing, it is likely that magistrates give some deference to the recommendation of a patient’s treating psychiatrist. Fitch, 14 N.C.Cent. L.J. 406, 455 (1984) (Winston-Salem study of involuntary commitments revealed that physician recommendations “generally control the outcomes of the cases”); accord Note, 62 N.C.L.Rev. 1158, 1163 (1984). But see Miller & Fiddleman, 60 N.C.L.Rev. 985, 1004-08 (questioning such deference).

There is an inherent conflict here between protecting society at large by confin ing the dangerous and protecting the individual patient by giving him the least restrictive environment. 7 Ideally, both goals would be met. In practice, however, they are not and probably never will be. Note, 63 N.C.L.Rev. 241, 255 (1984) (problems of overinclusiveness and underinclusiveness; present system does not even approach goal of perfect balance). Therefore, this court must attempt to decide the ramifications when they are not compatible, recognizing the potential to swerve too far to either side if it decides incorrectly.

The court finds both of the choices considered by courts thus far — either declining to recognize a duty at all, or recognizing a duty and liability under a traditional negligence theory — to be unsatisfactory. Instead, it believes a compromise is in order.

That compromise is a heightened standard of culpability required for liability. This has been done in the judicially-created “business judgment rule,” and the court believes that a similar standard of review is justified here.

In the business judgment rule, courts defer to the decisions of disinterested directors absent bad faith or self-interest. Many of the considerations cited as justifications for the business judgment rule are applicable to the present case. For example, as with business decisions, the court is not particularly qualified to review commitment decisions involving mental health and dangerousness. In addition, these types of commitment procedures require quick action, and “after-the-fact litigation is a most imperfect device to evaluate” those decisions, as in the corporate setting. Joy v. North, 692 F.2d 880, 886 (2d Cir.1982), cert. denied, 460 U.S. 1051, 103 S.Ct. 1498, 75 L.Ed.2d 930 (1983); see R. Robinson, North Carolina Corporation Law and Practice § 12-6 at n. 13 (Supp.1984) (citing Joy as “a good ... summary of the business judgment rule”). Finally, policy considerations favor giving psychotherapists, as well as corporate directors, significant discretion to use their best judgment, recognizing that “[a] rule which penalizes the choice of seemingly riskier alternatives ... may not be in the interest” of the parties or society. Joy, 692 F.2d at 886.

The court believes that a limited standard of review is the most appropriate to adopt here, where it is in society’s interest to review therapists’ decisions to some extent, but where it seems unfair to subject therapists’ immediate decisions to retrospective review by courts and juries who have far less expertise in an area in which the answers are never easy. Limited standards of review have been similarly adopted by courts in other areas in which deference to decisionmakers with greater expertise is desirable. See, e.g., Berry v. Bean, 796 F.2d 713 (4th Cir.1986) (court refused to look beyond face of military commanding officer’s decision to exclude civilians from base).

Under such a “psychotherapist judgment rule,” the court would not allow liability to be imposed on therapists for simple errors in judgment. Instead, the court would examine the “good faith, independence and thoroughness” of a psychotherapist’s decision not to commit a patient. Joy, 692 F.2d at 888. Factors in reviewing such good faith include the competence and training of the reviewing psychotherapists, whether the relevant documents and evidence were adequately, promptly and independently reviewed, whether the advice or opinion of another therapist was obtained, whether the evaluation was made in light of the proper legal standards for commitment, and whether other evidence of good faith exists.

In that such a rule provides a therapist with protection against civil liability for errors in judgment, it is consistent with North Carolina and federal law. North Carolina law provides a physician, hospital, or other official who “follows accepted professional judgment, practice and standards” with immunity in ’connection with the “examination, management, supervision, treatment, or release of a client.” N.C.Gen.Stat. § 122C-210.1 (Cum.Supp. 1985) (emphasis added). This immunity expressly extends not only to these responsibilities, but also to immunity “for actions of the client.” See also Pangburn, 326 S.E.2d at 371 (chilling effect of liability mitigated by qualified immunity). 8

It is also reasonable to conclude that the federal government, when it subjected itself under the Federal Tort Claims Act to the liabilities of state law, may have intended to get the benefits of analogous state immunities. It is doubtful that the federal government wanted its employees held to a higher standard than comparable state employees. Moreover, the federal government recently expressed itself in this area of the law. In the Protection and Advocacy for Mentally 111 Individuals Act of 1986, P.L. 99-319 (May 23, 1986), Congress stated that the Act’s expanded protections of the mentally ill nevertheless did not “obligate an individual mental health or health professional to administer treatment contrary to such professional’s clinical judgment.” Id. § 201(3)(B) (emphasis added). 9

The court thus finds that a “psychotherapist judgment rule” is the most appropriate in this area of the law. It gives society some check on the actions of therapists, ensuring that the interest of the public will be considered, but allows therapists to make these “delicate and difficult” decisions according to their more informed judgment, protecting the interests of their patients whose voices otherwise might go unheard.

C. Application of Standards to Facts

Having decided on the proper legal theory, the court now applies it to the facts to discern whether plaintiff’s case is sufficient to survive defendant’s motion for summary judgment. The court finds that it is not.

In a motion for summary judgment, the court construes all ambiguities in favor of the nonmoving party. Summary judgment should be denied if any genuine issue of material fact remains for trial. Fed.R. Civ.P. 56.

As noted earlier, whether defendant acted in good faith entails an examination of (1) defendant’s competence and training, (2) the adequacy, promptness and independence of defendant’s review of the patient’s case, (3) defendant’s efforts to check his judgment against the opinions of the other therapists, (4) defendant’s application of the proper standards, and (5) other evidence indicating the physician’s good or bad faith.

The first three factors tend to show good faith. Avery’s case was reviewed twice by a board made up of eight Duke University faculty members, seven of whom were board-certified in psychiatry. As for the sufficiency of the review, plaintiff makes no contention that the board did not act quickly or that they did not fully and independently review the records available to them.

Plaintiff argues that the fourth factor has not been met here, however. Plaintiff contends, and defendant does not deny, that all of the psychotherapists at issue believed that North Carolina law required that a patient be “psychotic,” or “out of touch with reality” before he could be involuntarily committed. Plaintiff has presented affidavits from doctors and other professionals that the “mental illness” required by the statute for involuntary commitment entails a variety of disorders other than “psychosis.” Without examining the intricacies of the various ailments in detail, the court agrees with plaintiff that the plain meaning of the statute would appear to be broader than the definition given by the psychotherapists in this case. N.C. Gen.Stat. § 1220-3(21) (Cum.Supp.1985) (statutory definition of mental illness); see Hiday, 60 N.C.L.Rev. 1027, 1028 & n. 8 (1982) (only some “mental illnesses” make a person incapable of rational decisions). As a result, this overly restrictive interpretation cuts against a finding that the therapists acted in good faith.

There is another factor present in this case, however, that convinces the court that the defendant’s good faith is beyond dispute. This factor is the realization that not only did Avery threaten IBM, but also he threatened the therapists themselves. These threats were serious and direct. Thus, it seems evident to this court that the therapists’ decision not to seek Avery’s commitment, a decision which involved a serious personal risk to themselves, was made in the utmost good faith.

What society in general, and North Carolina in particular, desires and expects psychotherapists to do when considering whether a patient should be involuntarily committed is to actively consider the public interest and to use their professional judgment in light of that interest. 10 The VA doctors have performed this duty in the instant case. In sum, the court does not believe that a jury in this case could reasonably find that the defendant’s decision not to commit Avery was not made in good faith. Accordingly, no material issue remains for trial, and summary judgment for defendant is appropriate.

D. Addendum

North Carolina does not have a certification statute. Such a statute would enable this court to enlist the assistance of the North Carolina courts in resolving difficult questions involving North Carolina law, such as those presented in this case. The court notes that the Tenth Circuit has taken advantage of certification statutes on two separate occasions to ask the relevant state courts how they would rule on these very issues. Durflinger, 727 F.2d 888 (10th Cir.1984); Brady, 751 F.2d 329 (10th Cir.1984).

E. Conclusion

For the reasons stated above, defendant’s motion for summary judgment will be granted.

1 . The Pangbum case held that a third party might recover for a psychotherapist’s wrongful release of a mental patient from a hospital, both because the plaintiffs complaint represented a valid claim and because any psychiatric immunity was not absolute. Bradley Center involved a psychotherapist’s failure to involuntarily detain a patient who had been voluntarily committed but who exercised his release option against medical advice. Georgia law allowed a 48-hour detainment in such circumstances.

2 . The court does not agree, for example, with Brady’s conclusion, quoted by other courts, e.g., Abernathy v. United States, 773 F.2d 184 at 190 (8th Cir.1985), that expanding a psychotherapist’s liability beyond “specific threats to specific victims” would "closely approximate a strict liability standard of care.” Brady v. Hopper, 570 F.Supp. 1333 at 1339 (D.C.Colo., 1983). Clearly a plaintiff still must prove wrongful conduct by a psychotherapist in order to recover, and a jury would be so instructed.

3 . Tarasoff, 131 Cal.Rptr. at 22, 551 P.2d at 342 ("We depart from this fundamental principle [of no affirmative duty] only upon the balancing of a number of considerations.").

4 . Commitment raises practical difficulties as well. In North Carolina, for example, there are an estimated 45,000 chronic mentally ill, while the state’s four mental hospitals house only about 3000 of the severely ill. At least one of these hospitals, Dorthea Dix, already faces overcrowding problems. Greensboro News & Record, Aug. 27, 1986, at B1 cols. 4-5.

5 . Several commentators have suggested that psychotherapists may have a duty to seek involuntary commitment of their dangerous patients. E.g., Note, 19 New Eng.L.Rev. 597, 616 (1984). Cf. Prosser & Keeton on Torts, § 33, at 203 & n. 99 ("unique position to prevent the harm").

6 . The only two empirical studies of this area of the law at least intimate that overcommitment is a possible effect of a mere duty to warn. Note, 1984 Wisc.L.Rev. 443, 478 ("some slight support" for idea that Tarasoff led to increased involuntary commitments, although improper commitments not examined); Comment, 31 Stan.L.Rev. 165, 188-89 & n. 124 (1978) (psychotherapists tended to overpredict dangerousness before Tarasoff; some responding to survey volunteered that they were more likely to commit after Tarasoff).

7 . The trend in recent years appears to have been toward restricting, rather than expanding, individual patient rights. The 1981 amendments to North Carolina’s Mental Health Act, for example, changed the standard for involuntary commitment from "imminently dangerous” to merely "dangerous.” This change may indicate an increased willingness by the North Carolina General Assembly to limit protections against overcommitment to those constitutionally required. See also Miller & Fiddleman, 60 N.C.L. Rev. 985, 1004-06, 1018-20 (1982) ("swing of pendulum from favoring release to favoring commitment”).

8 . In Pangburn, the North Carolina Court of Appeals construed a previous immunity statute as providing only qualified immunity to persons deciding whether or not to release mental patients. The court held that the statute did not "protect a tortfeasor from personal liability for gross negligence and intentional torts.” This standard is consistent with the "good faith” standard adopted by this court because any intentional wrongdoing or gross negligence by psychotherapists in their involuntary commitment decisions would probably constitute "bad faith."

9 . The psychotherapist judgment rule is consistent with opinions in the Fourth Circuit as well. Hasenei [v. United States, 541 F.Supp. 999 (1982) ] suggested a "competence” type of standard that supposedly would defer to valid psychiatric decisions. In a somewhat different context, the Fourth Circuit held in Thomas S. v. Morrow, 781 F.2d 367 (4th Cir.), cert. denied, — U.S. -, 106 S.Ct. 1992, 90 L.Ed.2d 673 (1986), that a court reviewing North Carolina's mental health training or treatment plans was “required" to accept the plans recommended by North Carolina's professionals unless important relevant factors were not considered or the recommendation represented a "substantial departure from accepted professional judgment or standards.” Id. at 375.

10 . The court does not decide whether the psychotherapist judgment rule would be appropriate in a duty to warn case. Less deference to psychotherapists' individual judgments in the warning area may be justified because the burden placed on the psychotherapist and the countervailing policy considerations are not as great. See Note, 29 Hastings L.J. 179, 207 (1977) (advocating duty to warn broader than commitment standards). In other words, the fact that a psychotherapist’s decision not to seek commitment of his patient was made in good faith does not mean that he has fulfilled any duty to warn that may have attached.