Newletter Masthead
April 2001· Vol. 26, No. 2, pp. 13-17

Williamson v. Liptzin appeal

Issues of liability for a patient's unexpected act of violence

Paul B. Herbert JD, MD

In October 1998, a North Carolina jury made national headlines by handing down a $500,000 malpractice verdict against a university health service psychiatrist and in favor of a former patient who went on an unprovoked and random shooting spree, killing two people and wounding several others (and who was himself wounded by responding police and subsequently tried for murder). On December 19, 2000, the verdict was overturned on appeal and the claim dismissed. Williamson v. Liptzin, 2000 N.C. App. LEXIS 1276 (N.C. Ct. App. Dec. 19, 2000).

The Facts

In May 1990, while an undergraduate at the University of North Carolina, the plaintiff visited the university health service as a "walk-in" and was counseled for relationship issues and academic problems. The progress note concluded that the plaintiff's problems were "fairly normative" and no follow-up was scheduled.

Two and one-half years later, in September 1992, as a 24-year-old law student on the same campus, the plaintiff began screaming at other students and inexplicably striking himself in the face. He was referred to the university health service and from there was admitted involuntarily to the university hospital. He disclosed to his inpatient treaters that he had been hearing a voice for eight months and that he believed he was telepathic; he also revealed that he had a gun in his apartment. His stay was brief, as he successfully challenged his commitment; the judge ordered the plaintiff discharged "and recommended that patient seek out-patient psychiatric counseling." Hospital staff arranged for the plaintiff's parents to confiscate the gun; the discharge diagnosis was "rule-out schizophrenia." The plaintiff evidently did not pursue follow-up.

Six months later, in early March 1994, the plaintiff was again referred to the university health service "after he disrupted class at the law school by announcing that he was a 'telepath.'" Intake staff diagnosed the plaintiff with "rule-out schizophrenia" and recommended outpatient treatment, including medication, which the plaintiff refused; the staff concluded that involuntary hospitalization was "not appropriate as student denies danger to self or others."

A law school dean informed the plaintiff that he might not be recommended as a candidate for admission to the bar unless he received counseling; the plaintiff promptly returned to the university health service. He saw the defendant for six counseling sessions (ranging from twenty to sixty minutes) over a ten-week period commencing on March 8, 1994.

At the first session, the plaintiff reported that he had believed he was a "telepath" for two years and acknowledged that he drank a six-pack of beer daily and used marijuana occasionally. Notwithstanding having reviewed the intake form from the plaintiff's May 1990 health service visit, the discharge summary from the September 1992 hospital stay (documenting eight continuous months of auditory hallucinations) and the March 1994 intake paperwork, the last two diagnosing "rule-out schizophrenia," the defendant "decided to record the more 'generous' diagnosis . of 'delusional disorder grandiose' . so as not to deprive plaintiff of the opportunity to practice law."

The defendant prescribed Navane, with which the plaintiff was substantially compliant during his ten weeks under the defendant's care. At the fourth of the six sessions, on April 5, 1994, defendant apprised the plaintiff that he would be retiring in June and suggested that the plaintiff "consider the possibility of seeing somebody on a regular basis in therapy, and that [defendant] would be happy to make a referral for him; it would probably make sense to do this sooner rather than later."

At the final session, on May 25, 1994, the plaintiff still did not know whether he would be staying in Chapel Hill for the summer or going home to stay with his parents. The defendant recorded in the chart that the plaintiff knew he was leaving the health service, that the plaintiff would be seeing his replacement in the fall semester and that the plaintiff knew he needed to contact the defendant's replacement to have his Navane prescription refilled. The defendant furnished a prescription for thirty Navane (about two weeks' worth, assuming b.i.d. dosing) and charted that the plaintiff was "content to stay on [Navane]." The defendant instructed the plaintiff to make an appointment with the defendant's replacement at the university health service or see his family doctor or the local community health service if he elected to spend the summer at home with his parents.

Under the defendant's care, on Navane, the plaintiff dramatically improved: the hallucinations and the telepathy delusions remitted, the alcohol and marijuana use diminished (though continued), the plaintiff completed his Spring semester law classes without incident, improved his grades and participated in a writing competition. Friends reported that he was "more 'like his old self.'" The plaintiff had no thoughts of violence or self-harm (and no history of violence, by the way, other than the peculiar episode of striking himself in the face twenty months earlier, in September 1992, which led to his sole hospitalization).

After termination with the defendant in May 1994, the plaintiff returned home to his parents' house for the summer. He stopped taking his antipsychotic medication and had no further contact with any mental health professional, yet continued to function well for over six months. The appellate court tersely reports:

"Plaintiff returned to Chapel Hill in August of 1994 for the fall semester. He attended virtually all of his classes and did not disrupt any of them. He passed all of his courses, managed his finances, and took care of his day-to-day needs, such as grooming, eating, and shopping. He took trips alone in his car, including trips to Connecticut and New York City over Christmas break."

The court seems to think that the last sentence in the foregoing quote bolsters the previous sentences, evidencing that the plaintiff's inner world remained integrated. A mental health professional would, of course, take at least a more skeptical, if not contrary, view of the plaintiff's taking "[long] trips alone in his car . over Christmas break," when suddenly stripped of the emotional support, external structure and continuity of living with his parents over the summer and then returning to the law school community in the fall.

It seems likely that, off medications for about seven months, the plaintiff's latest psychotic decompensation had begun by December 1994. Soon calamity struck:

"In January 1995, plaintiff returned to Chapel Hill and began living out of his car. He stopped attending classes and purchased guns and ammunition. In addition, the plaintiff returned to his parents' home, to retrieve a [sic] M-1 rifle, the gun UNC hospital staff noted he possessed. The weapon had been [at his parents'] since plaintiff's hospital stay in 1992. On 26 January 1995, eight months after his last session with defendant, plaintiff randomly fired the M-1 rifle at unarmed people in downtown Chapel Hill, killing two of them."

The Trial

The plaintiff's malpractice suit focused mainly on the defendant's misdiagnosis of "delusional disorder" and on the rather inchoate follow-up arrangements.

The "delusional disorder" diagnosis opened the door for the plaintiff to claim, plausibly, that defendant misled him about the seriousness of the illness and even gave him "permission," in a sense, to cease taking his antipsychotic medication:

"Plaintiff believed that his mental illness was temporary and that the medication was a short-term measure. According to plaintiff, defendant told him 'that in his opinion, [plaintiff] was probably not really schizophrenic or psychotic.' Plaintiff further stated that defendant told him that 'if someday [he] wanted to go off the medication, that [he] could do that if [he] told someone [he] trusted." [] Plaintiff spent the summer at his parents' home .. He did not visit the community health center or Student Services. Plaintiff decided to stop taking Navane for a few days . [and] felt physically better. He determined that he would stop taking his medication indefinitely and informed his parents of that decision."

The plaintiff's three experts, two psychiatrists and a psychologist, hammered home this point. They concurred that the correct diagnosis was unquestionably schizophrenia, quite different from (and more ominous than) delusional disorder, that the advice that the plaintiff could go off his medication, if given, "would have been improper and an 'invitation to not comply with the recommended therapy,'" and that psychotic decompensation would then be inevitable, with the downward spiral fueled by deteriorating insight and judgment and continuation of substance abuse. The testimony of Dr. Warren, the psychologist, must have been especially weighty for the jury:

"Dr. Warren stated that plaintiff was not competent to take charge of his medical treatment at the time his therapy with defendant ended. Dr. Warren testified that there's nothing in the record that suggests that [plaintiff] got that [sic] information that he needed in order to make decisions about whether or not he had a major mental illness, whether or not he needed to take medication on a long-term basis, what he needed to do in case the symptoms got worse. [] Plaintiff reported to Dr. Warren on the day following the shootings that defendant had advised him that he could discontinue his medication if he told someone he trusted. [] . Dr. Warren believed that because plaintiff did not understand the seriousness of his illness, he could not make competent decisions concerning treatment."

Some additional, less important, points were made by the plaintiff's two psychiatrist experts. Both stressed the plaintiff's "history of noncompliance" as a particular warning sign that the defendant should have heeded. This does not seem especially fair, since a lay jury cannot be expected to know that "noncompliance" is so common in the initial stage of a major mental illness as to be virtually part of the disease itself. One of the psychiatrists further testified that the defendant's failure to go behind the plaintiff's discharge summary and read the records (viz., progress notes) of the plaintiff's hospitalization fell below the standard of care. This is arguable, perhaps, as the defendant's clinic and the hospital were part of an integrated (university) health care system but it seems somewhat "gilding the lily" insofar as the discharge summary itself clearly states "rule-out schizophrenia" as the diagnosis. The testifying psychiatrist does not seem to have identified any additional data, appearing only in the progress notes, that was missed.

Very significantly, since this was the issue on which the appellate court was later able to rescue the defendant from the jury's verdict, all three of the plaintiff's experts acknowledged on cross-examination that, notwithstanding the inevitability of the plaintiff's psychotic decompensation off medication, with possible risk to his own safety, his becoming dangerous to others was not foreseeable. As one expert put it, "nobody's crystal ball is that good . [even] if defendant had given plaintiff the name of a specific doctor to visit during the summer of 1994, [the expert] still could not predict [sic] what would have happened."

The defendants' three psychiatrist experts all made essentially two points. First, psychosis does not equate to dangerousness and dangerousness is notoriously unpredictable; as one put it, "because the base rate of violence is so low, and most schizophrenics aren't violent and most normal people aren't violent either, that demographic data does not get you anywhere in predicting dangerousness."

And second, university mental health is inherently "short-term . before [the students] have to make plans with or without [the psychiatrists'] assistance to get further treatment, or at least make choices in their life." This second point seems both factually dubious and of, at best, murky relevance. University students are a fairly stable catchment, often followed for several years (the plaintiff's contacts with his campus mental health service spanned four years, from May 1990 to May 1994); it is not clear that community mental health center clientele or private outpatients, given the limitations of health insurance coverage, have characteristically lengthier or deeper courses of treatment nowadays. And, in any event, even the three defense experts made no effort, so far as the appellate opinion indicates, to defend either the diagnosis in this case or the lack of specific follow-up arrangements, professional functions from which university psychiatrists surely are not uniquely exempted.

The jury found the defendant to have been negligent in his care of the plaintiff which in turn resulted in the shooting spree and led to the plaintiff's injuries: pain from being shot in the legs by the police (acting to stop the shooting spree), consequent surgery for the leg wounds, the distress of suffering through a murder trial, the ensuing psychiatric hospitalization following the criminal trial jury's insanity acquittal and "not [being] able to continue his legal studies or pursue a possible career." The jury found no contributory negligence by the plaintiff. The award was $500,000.

The Appellate Decision

The North Carolina Court of Appeals reversed. Tellingly, the defendant appealed mainly on grounds of lack of "proximate cause," and the court limited its focus to that issue, surgically excising the issue of negligence (from this negligence case).

Not once in its analysis alluding to the defendant's misdiagnosis or the scanty follow-up arrangements (or any action of defendant), the appellate court, citing first-year law school chestnuts such as Palsgraf v. Long Island Railway and Prosser on Torts, observed:

"The element of foreseeability is a requisite of proximate cause .. '[I]t is inconceivable that any defendant should be held liable to infinity for all the consequences which flow from his act,' [thus] some boundary must be set."

Recognizing that this rather mystical concept of how far downstream responsibility for one's negligence extends is almost always reposed in the jury (under appropriate instructions), the court nonetheless declared that this was that extremely rare situation where a court must step in and find a lack of proximity in causation as a matter of law:

"Despite [the] attenuated chain of events, plaintiff contends that the testimony of his experts was tantamount to the issue of foreseeability and was more than sufficient to establish that 'some' injury was foreseeable ... [W]e cannot agree . [T]he expert testimony presented by plaintiff established what was merely possible and not what was reasonably foreseeable."

The court stressed the long eight months between termination of treatment and the rampage:

"[P]laintiff's injuries were too remote in time, and the chain of events which lead to plaintiff's injuries was too attenuated for defendant's actions to be the proximate cause of plaintiff's injuries. It was eight months between plaintiff's last visit with defendant and the incident which led to his injuries. Plaintiff was, by all accounts, functioning normally when he last visited defendant in May 1994 . In August 1994, plaintiff returned to law school and began his fall classes . [H]e attended virtually all of his classes during the fall semester, without disruption, and passed every course . In January 1995, . [e]ight months after his last visit with defendant, plaintiff shot and killed two individuals in Chapel Hill, despite never expressing any intent to do so. Defendant simply could not have foreseen that as a result of this attenuated chain of events, eight months after his last appointment, plaintiff, who expressed no violent intentions or threats, would be wounded during a shootout, tried for capital murder, committed to a mental institution, and not able to continue his legal studies or pursue a possible career."

Seemingly even more important to the court was the intrinsic slipperiness of dangerousness predictions (as argued by the American Psychiatric Association, and allied amici). For this point, dictum from Tarasoff v. Regents of Univ. of California was offered up: "We recognize the difficulty that a therapist encounters in attempting to forecast whether a patient presents a serious danger of violence." The Tarasoff citation is doubly ironic insofar as (a) the language quoted cuts against the very holding of Tarasoff (viz., that a psychiatrist can be held liable for failing to warn of a patient's future violence) and (b) North Carolina remains among the roughly one-third of the states that have not addressed the Tarasoff question, pro or con, by statute or caselaw.

In this case, the court persuasively observed:

".North Carolina courts are reluctant to hold a person liable where the chain of events which led to the resulting injuries is unforeseeable, remote, and attenuated, even though 'some' injury to plaintiff was 'possible.' . The contemplation of what 'might' happen . and the consideration of 'risk factors' for violence to oneself which may or may not lead to a risk of violence to others, is simply not sufficient as a matter of law to establish the foreseeability of plaintiff's injuries or the circumstances in which those injuries arose. Furthermore, evidence of 'risk factors' for potential violence, such as gun ownership, being under a certain age, or being of a certain gender, implicates a large portion of the population and is simply insufficient in and of itself to prove foreseeability. Given the lack of evidence of violence or any threats of violence on plaintiff's behalf [sic], the connection between negligence and the injury appears unnatural, unreasonable, and improbable.' . We therefore conclude that the expert testimony presented by plaintiff established what was merely possible and not what was reasonably foreseeable."

The court's reasoning is labored and unconvincing as it struggles to justify preempting the jury on quintessentially a fact question and to wrestle with inconvenient North Carolina precedents. It is difficult to share the court's view, for example, that the defendant in the seminal state case on foreseeability, Hairston v. Alexander Tank & Equipment Co., had more reason to foresee harm in that case and was therefore properly held liable. A new car was sold, on which one of the wheels had an untightened lug nut; the wheel flew off but plaintiff safely stopped the car and was off the highway surface replacing the wheel when another driver, with supervening negligence, hit and killed the plaintiff. (One cannot help suspecting that the real distinction between the cases was the disparity in charm of the respective plaintiffs, a widow suing for loss of her husband's life versus a killer suing for loss of his prospective legal career.)

The point would seem to be that the court, cognizant of the heuristic and policy ramifications of its decision in the sensitive and important realm of mental health, was determined to reach a sound result, despite a rather weak hand dealt it by virtue of defendant's conduct.

"Imposing liability on a psychiatrist in an outpatient, short-term care setting for the actions of a patient that were at most based on risk factors and not foreseeability . would encourage psychiatrists and other mental health providers to return to paternalistic practices, such as involuntary commitment, to protect themselves against possible medical malpractice liability . North Carolina's policy on the mentally ill promotes less restrictive methods of treatment and more patient autonomy . 'a patient has an active responsibility for his own care and well-being' . It would therefore be irrational to promote unnecessary, more restrictive practices in affirming the judgment below."

While it is axiomatic that trial courts are concerned with justice (retail jurisprudence) whereas appellate courts address only questions of law (wholesale jurisprudence), it occurs not infrequently that appellate courts cannot help but feel responsible to rectify outcomes too feckless or unconscionable to bear, even where no legal principle has been clearly traduced. When they do so, the counter-axiom comes to the fore: "hard cases make bad law." Recognizing the dubiety of encouraging fact-challenges to tort verdicts, the court thus is at pains to curtail the precedential significance of its holding:

"We recognize that [North Carolina] jurisprudence in the area of proximate cause is quite varied . We further recognize that it is only in the rarest of cases that our appellate courts find proximate cause is lacking as a matter of law .. This is one of those rare cases. We conclude that given the very specific and novel factual scenario presented by this case, defendant's alleged negligence was not the proximate cause of plaintiff's injuries."


As legal precedent, there is considerably less to Williamson than meets the eye. In the first place, as a state court decision the case constitutes law only in North Carolina and, at that, carries less weight as an intermediate appeals court decision than would a decision of the state's supreme court. (The state supreme court could docket the case for review but this seems not a particularly enticing thicket.) Beyond North Carolina, the decision, not being law, would have weight only as "persuasive authority." It will have little, for two reasons.

First, the court has deliberately made its holding as narrow as possible, emphasizing that proximate cause determinations are necessarily sui generis (unique to each case). To underscore the point, the court recites over and over the peculiar facts of this case, including the eight months delay (during most of which the plaintiff was outwardly high-functioning), the absence of any history or threats of violence, the brief course of treatment and the setting (a university health service) and so on.

Second, in truth, the case was about contributory negligence, not proximate cause:

"Plaintiff was, by all accounts functioning normally when he last visited defendant in May 1994. Plaintiff spent the summer with his parents ., at which time he discontinued his medication and failed to visit a mental health center or to have his prescriptions [sic] refilled. In August 1994, plaintiff returned to law school and began his fall classes. Plaintiff testified that his hallucinations began to resurface gradually and achieved fruition sometime in August or September. However, the plaintiff attended virtually all of his classes during the fall semester, without disruption, and passed every course. He maintained his daily needs, including eating, grooming, shopping, and managing his financial affairs. Furthermore, after completing the semester, the plaintiff took two long trips alone, after which time he returned to his parents' home .. In January 1995, plaintiff returned to Chapel Hill. Only at this time did plaintiff begin living out of his car, stop attending classes, and purchase guns and ammunition . [and three weeks later] shot and killed two individuals ."

The picture is of a rational individual, capable of carrying through on very demanding and complex tasks (e.g., successfully navigating a law school semester and properly maintaining his financial affairs), who deliberately chose for an eight-month period not to attend to his mental health, notwithstanding the resurfacing of auditory hallucinations that he knew in the past had augured psychological disintegration. The result: plaintiff ended up killing two complete strangers - and was acquitted (on grounds of insanity).

In such circumstances, one somewhat marvels at the plaintiff's hubris in suing his former psychiatrist for his (the plaintiff's) injuries and one is left agape at the jury's decision that it was all the psychiatrist's fault.

However, appellate courts cannot (for all practical purposes) touch a negligence (or contributory negligence) finding, the paradigmatic jury issue. Hence the court's opinion, teetering instead on the rickety foundation of "proximate cause."

As clinical precedent, however, Williamson is signal. Stripped bare, the facts are that defendant (apparently purposely) misdiagnosed schizophrenia as "delusional disorder," neglected to diagnose or target for specific treatment (such as AA meetings or partial hospital therapy) substance abuse in a six-pack-per-day (plus "occasional" marijuana) gun-owning schizophrenic, and made no specific follow-up arrangements at termination (occasioned by defendant's retirement).

But for a quite striking confluence of serendipitous facts, the appellate court could not have saved the defendant (and surely would not have been inclined to do so): the plaintiff harmed others rather than himself (the court views self-harm in this case as significantly more foreseeable); the plaintiff was extraordinarily high-functioning (attenuating the nexus between defendant's actions and the plaintiff's decompensation); a full eight months went by; the plaintiff made no attempt to pursue follow-up (as distinct from a mentally disturbed patient who might try on his own to make an appointment not arranged for him but fails); there were no documented threats or past acts whatsoever of violence (quite peculiar in a case that eventuates in a shooting spree); and the clinical setting was a university health service (which the court implies unconvincingly carries a lesser standard of care with respect to diagnosis and follow-up).

Further, the defendant chanced on a court unusually determined to undo a jury fumble, and remarkably agile in doing so. The court's non-discussion of the defendant's non-diagnosis (and non-referral) of the plaintiff's alcoholism is eloquently (albeit silently) emblematic. The defendant, in short, was luckier in dodging the plaintiff's legal bullets than were the January 1995 victims against his actual bullets.

This clinical lessons would seem clear. First, do not diagnose falsely, regardless of how benign the motive. A cardiologist would not deliberately overlook a basketball prospect's serious valve pathology and then expect to be exonerated in a wrongful death suit by asserting that he had not wanted to stand in the way of the player's athletic career. Second, where follow-up is clearly indicated (as with a young and noncompliant substance-abusing schizophrenic), more must be done (and documented) than urging the patient himself to make appropriate arrangements.

Paul Herbert is a clinical instructor in psychiatry at Yale University.