The Case That Has Psychiatrists Running Scared: Ahmed v. Stefaniu ================================================================= * Dylan R. Glancy * Graham D. Glancy ## Abstract In the three decades since the definitive decisions regarding duty to warn and protect, several cases have shaped psychiatric practice. In this article we present a tragic case that we postulate may have a significant effect on psychiatric practice in Canada. A psychiatric patient murdered a relative some 50 days after the patient left the hospital. On the day before his release, the patient's status was changed from involuntary to voluntary. We argue that this case may result in a reversal of current mental health and social policy, whereby psychiatrists will be less willing to release difficult and potentially dangerous patients. To provide context to the case, we review some concepts of the literature and recent legal cases pertaining to the release of such patients. Psychiatrists have long been in the position of having to balance the rights of patients against protecting the public, knowing that failure on the latter score could raise a legal question of professional negligence. Canadian law requires four essential components to prove professional negligence.1 First, the defendant must owe the plaintiff a duty of care. Second, it must be proven that the defendant breached this standard of care. Third, it must be shown that the plaintiff has suffered an injury or loss. Finally, the court must find that the defendant's conduct has been the actual and legal cause of the plaintiff's injury. The nature of a duty to protect the public has been argued in psychiatric and legal circles for some time. The definitive discussion begins with the *Tarasoff* 1 and 2 decisions in 19742 and 1976.3 In *Tarasoff* 1,2 it became law that therapists had a duty to warn potential victims of violence from foreseeably dangerous patients. In *Tarasoff* 2,3 the therapist is mandated to use reasonable care to warn and protect an intended victim from danger. Many states in the United States have adopted *Tarasoff* statutes as well as *Tarasoff*-limiting statutes.4 Canada has been a little slower to come to conclusions about duty to protect. No doubt we have had the benefit of watching 20 years of U.S. experience. However, in a 1991 Canadian case, *Wenden v. Trickha*,5 a seriously ill psychiatric patient escaped from the hospital and stole a car. He drove erratically at high speed, acting on a delusion that the car was a time machine. He drove through a red light, crashed into another car and seriously injured Ms. Wenden. The court ruled that the hospital may have been liable if it had been possible to identify a victim, but concluded that it was not possible in this particular case. However, this case opened the door to future litigation when escaped patients cause injury to another. In a later case, *Smith v. Jones*,6 a psychiatrist assessed a client who had been referred by a lawyer for a defense psycholegal assessment. Mr. Jones (not his real name) told the psychiatrist that he had elaborate plans to kidnap a prostitute, and then torture and murder her. The doctor called the client's lawyer and informed him of his findings. The lawyer told the doctor that he did not intend to use any report in court and that the communications between the doctor and the client were to be kept confidential. The doctor called his own lawyer, retaining counsel, who took the unusual step of applying to the court to breach solicitor-client confidentiality. The case was fast-tracked to the Supreme Court of Canada, which concluded that if a patient or client makes a serious threat of imminent harm to a third party, then the doctor is obliged to take steps to warn or protect the third party. The Court noted that the threat must be imminent and pose serious bodily harm to an identifiable third person or group. The Court declined to outline the steps necessary to warn or protect potential victims, leaving such details to assessment on an individual case basis. It should be noted that this case applied to solicitor-client privilege, the highest privilege in the land,7 and led to the development of a position paper on the duty to protect by the Canadian Psychiatric Association (CPA).8 The CPA essentially adopted findings from a group of authors9 who published an algorithm on the issue to guide the clinician through this difficult process. We present a case, *Ahmed v. Stefaniu*,10 that put the psychiatrist's duty to warn and protect at issue. The case was taken to the provincial Court of Appeal, which is one level down from the Supreme Court of Canada in importance. Leave to appeal to the Supreme Court was sought, but denied. As the reader will see, the case has troubling social policy implications. ## The Facts William Johannes had been an inpatient at a different hospital from the one where Dr. Stefaniu had privileges to admit patients. In the summer of 1996, Mr. Johannes' condition deteriorated. He was reported to be acting bizarrely, with aggressive and paranoid ideas. On September 25, 1996, his sister, Roslyn Knipe, with whom he resided, contacted her family practitioner and informed him that Mr. Johannes had threatened to hurt her if she did not prove that she was “on his side” within two weeks. The police were informed and Mr. Johannes was taken to the hospital and admitted involuntarily. He was reported to have severe paranoia, with lack of insight and a potential for violence. The physician found that he was likely to cause serious bodily harm to another person, and he was also declared incapable of consenting to treatment. Mr. Johannes appealed his involuntary hospitalization to the Consent and Capacity Review Board (CCRB), which concluded: The board is of the opinion that at this time, without treatment, there is a likelihood that the patient if he left hospital would continue to deteriorate to the point where the likelihood exists that he will cause serious bodily harm to another person. He would simply lose control due to anger and frustration [Ref. 10, p 3]. In mid-October, the doctor noted that Mr. Johannes had deteriorated further, with the potential for self harm and harassing others. On October 24, 1996, he struck another patient and had to be placed in two-point restraints due to his level of agitation. On October 31, 1996, he threatened the Chief of Psychiatry and another psychiatrist. Following this episode, he was involved in a fight with staff, and security guards were called to subdue him and place him in four-point restraints. It is recorded that during his stay in the hospital, physical restraints were used on Mr. Johannes on 25 occasions. In mid-November, security records noted that he attempted to attack two patients. On December 2, 1996, the doctor assessed him and noted that he remained delusional and paranoid. A progress note on December 3, 1996, described him as very angry, loud, and intrusive, with threatening body language in a rigid posture. He was further described as extremely hostile, and on December 4, 1996, he threatened a nurse. It is reported that he became known as an extremely difficult patient.11 He wandered around naked and habitually sat next to the nursing station preaching loudly. At one point he incited a rebellion, encouraging other patients to refuse their medication. The staff was anxious to discharge him, some believing that he had a manipulative personality disorder. Several consultations were arranged, suggesting a division of opinion regarding Mr. Johannes' diagnosis. At least two of the consultants suggested he was faking some of his symptoms and acting out in a manner suggestive of a personality disorder. Nevertheless, the general impression over the course of his hospital stay was that he became less threatening and that there was a general trend of improvement in his behavior. On the evening of December 4, 1996, Stefaniu assessed Mr. Johannes after he had threatened a nurse and described him as having no signs or symptoms of paranoia or psychosis. She found him very appropriate, cooperative, and displaying a great sense of humor. Mr. Johannes told the doctor that he had no intention of harming himself or anybody else, including his sister. He informed the doctor that all of his behavior at the hospital had been staged and planned. Stefaniu noted this and a confession that he had faked his illness, although she later testified that she did not believe him. She concluded that he probably did not meet the criteria for an involuntary patient at that time and decided to meet with him the next day for further discussion. Following that meeting, she wrote the following progress note: Patient seen again today. Pleased he is finally released. Good mood, pleasant, cooperative, but inappropriately flirtatious (jokingly invited me for dinner). Has plans about job, picking up his car, and restarting life. No signs of delusions, paranoid thinking or psychosis. Denied any suicidal or/and homicidal ideas … [illegible] or plans. Alert and oriented—however remains provocative, Macho and in [*sic*] the same time angry and fragile [Ref. 10, p 4]. Acting on her opinion that Mr. Johannes no longer met the criteria to be detained in the hospital as an involuntary patient, Stefaniu changed his status from involuntary to voluntary under the Mental Health Act. She testified that she had considered factors that included the patient's general trend of improvement, his stated intention that he did not plan to harm himself or others, his response to medications, the decision of the CCRB, his previous admission to hospital, consultations with other psychiatrists,11 and conversations with his employer. She suggested that he remain in hospital as a voluntary patient, but he refused to do so. He also declined to continue outpatient care with his previous outpatient psychiatrist. He therefore was discharged from the hospital and moved back into his sister's apartment, where he resided with her and her two small daughters. The day after his release, the mental health director sent e-mail messages to all of the psychiatrists advising them that if Mr. Johannes returned to the hospital they should make every effort to transfer him to another facility. He subsequently returned to the hospital on more than one occasion to visit a female patient. It was recorded that such visits were disruptive and he was asked to leave the premises. On one occasion, he had to be escorted off the premises by the police and spent the night in jail. On January 21, 1997, approximately 47 days after his discharge, Mr. Johannes attended a different emergency department and was assessed by a psychiatrist. He was noted to be depressed but well dressed and articulate. He said that he was not capable of hurting himself or others. The doctor saw no indication of violence or potential for violence. He concluded that Mr. Johannes was not a danger to himself or to others and therefore did not meet the criteria for involuntary admission. The next day, Mr. Johannes attended yet another hospital emergency department where he was seen by the emergency department resident. She reviewed his medical history, including a history of his admission to the original hospital. He requested an immediate psychiatric consultation, but the doctor declined to order a consultation because she did not regard him as a danger to himself or others. She assessed him to be stable and suggested outpatient care. She consulted her emergency room supervisor and concluded that Mr. Johannes did not meet the criteria for involuntary admission under the Mental Health Act.12 No issue was taken at trial as to the standard of these assessments. On January 24, 1997, Mr. Johannes killed his sister in her apartment. At the time, he was said to be in a floridly psychotic, acutely delusional rage in which he believed that his sister was possessed by the devil. He was subsequently found not criminally responsible due to mental disorder (NCR-MD) and ordered to a maximum-security psychiatric hospital. His brother-in-law, Mr. Ahmed, and Ahmed's two daughters subsequently commenced an action against Stefaniu and others, alleging professional negligence. The plaintiffs called two experts, both of whom had extensive experience in general hospital psychiatry. They testified that Stefaniu failed to meet the standard of care expected of a psychiatrist in the circumstances. The defense called two experts who testified that the doctor acted in an honest and intelligent manner and that she met the standard required in her management of the patient. The jury accepted the evidence of the plaintiffs' experts in reaching its verdict, finding Stefaniu negligent in that she failed to meet the standard of care of a psychiatrist practicing in a general inpatient psychiatric unit in a community hospital when she made the decision to change Mr. Johannes' status to that of a voluntary patient. The court awarded the plaintiffs damages in the amount of $172,000. ## The Appeal The verdict was appealed on the following four grounds: the trial judge failed to instruct the jury properly on the law regarding the “honest and intelligent exercise of judgment” by a physician and failed to characterize properly certain admissions made by the plaintiffs' expert in his evidence; the trial judge failed to instruct the jury properly on the law regarding clinical judgment that is supported by a “reputable body of opinion within the profession” and failed to characterize properly certain admissions made by the plaintiffs' expert in his evidence; the trial judge failed to instruct the jury properly on the issue of causation, especially in view of the two emergency room assessments that intervened; the trial judge did not take into account that policy decisions require balancing the need for the protection of the public against the right of individual patients. The Court of Appeal stated, with no context, that “it should be noted that ′duty of care' was not raised as an issue on the appeal” (Ref. 10, p 5). ## The Decision The Court of Appeal dismissed the appeal. They stated that in their opinion, the trial judge had outlined the expert evidence in respect of the exercise of honest and intelligent judgment. They said that he did this in the context of all of the evidence. They stated that it was not material that he did not inform the jury that one of the experts for the plaintiffs conceded that the doctor may have made an honest error. They accepted that the trial judge's charge to the jury on causation was fair and adequate. They concluded that the verdict of the jury was not unreasonable and therefore it was open to the jury to find that, but for the change in Mr. Johannes' status to a voluntary patient, he would not have murdered his sister. The defense sought leave to appeal to the Supreme Court of Canada. The threshold to seek leave is high: the case has to be considered a matter of national importance. The defense argued that this was of national importance since it dealt with the liberty of subjects versus the protection of the public. These matters open up questions of social and public policy across the nation. However, the Supreme Court turned down leave to appeal, and as is their usual procedure, did not give any reasons. ## Analysis *Ahmed v. Stefaniu* was unusual in that complex medical cases are usually tried by judge alone in Canada. It was heard by a jury for complicated legal technical reasons. The jury found that Stefaniu had been negligent and did not meet the standard of care for a psychiatrist in a district general hospital. They found that the defendant's conduct was the actual and legal cause of the injury, pain, and suffering of the plaintiffs that resulted from the death of Ms. Knipe, their wife and mother. One of the potential points that the trial turned on was whether Stefaniu considered the appropriate factors when terminating Mr. Johannes' involuntary status. One of the plaintiffs' experts relied heavily on the decision of the CCRB. The Board heard their evidence on October 8, 1996, some three and a half months before the tragic incident—a long time in the context of the average stay in a district general hospital psychiatric unit. The patient's condition would have been expected to change drastically during that period, although in the actual case, the evidence was somewhat equivocal. The defense argued that such an error in judgment would not lead to a finding of negligence. The Court of Appeal believed that the jury had considered this evidence and noted, “There is often a fine line between error in judgment and failure to meet the standard of care… . [T] he jury could find that the line was crossed … [and] therefore the verdict was not unreasonable” (Ref. 10, p 8). In what it believed to be a strong point on appeal, the defense argued that even if Stefaniu's decision fell below the standard of care, two subsequent assessments demonstrated that Mr. Johannes' mental state had changed anyway, in that the patient did not meet the criteria for involuntary admission. Both of these assessments had occurred some weeks after discharge, but only days before the killing. The defense argued that this broke the chain of causation. Causation is a complicated concept, and oceans of ink and tears have been spilled attempting to come to a clear definition. The Supreme Court ruled in the case of *Snell v. Farrell*,13 that the legal or ultimate burden of proof is left with the plaintiff regarding causation. However, in the same case, the Court went on to say that in the absence of evidence to the contrary, an inference may be drawn although positive or scientific proof may not have been adduced. In a later case, *Athey v. Leonate*,14 the Court allowed that, on that occasion, the “but for” test could be unworkable and therefore it is only necessary to prove that the defendant's negligence materially contributed to the plaintiff's damages. The Court further noted that it is not necessary to establish that the conduct of the defendant is the sole cause of the plaintiff's injury. A recent case encouraged the Court to take a robust and pragmatic approach to analyzing the evidence on causation.15 Pengelley and O'Connor16 argue that hindsight bias is an important factor when the jury is considering whether a physician meets the required standard of care. They use evidence from the psychological literature demonstrating that learning of an outcome experimentally creates hindsight bias that cannot be disregarded. Even educating individuals about this and warning them not to be influenced by certain information has no effect. We see this as a particular problem in cases such as *Ahmed v. Stefaniu*. In that case, the terrible, tragic death of a mother was accentuated by the appearance of her children, both of whom gave evidence. It has been suggested that Stefaniu appeared cold, clinical, and severe on the witness stand. She was said to have appeared defensive.11 In contradistinction, the two daughters of the deceased instantly captivated the jury. They described the death of their mother with a grave demeanor and emotional restraint. Within a minute of the beginning of that testimony three jurors were weeping.11 In many civil trials, evidence of damages is presented during the trial even before a verdict has been reached. Therefore, the jury may be subjected to emotional information about the tragic effects of events on people. This is likely to influence their decision-making on the issue of causation. Pengelley and O'Connor16 note that bifurcated trials may be helpful. The first part of the trial can deal with evidence about the defendant's conduct, leading to a decision on the requisite standard of care, causation, and negligence. The next phase of the trial, perhaps involving a different trier of fact could then hear evidence on the damages issue. At present, Canadian civil courts rarely use this process, but it is not unknown. The argument against it is that the logistical problems would further burden the cumbersome legal system and significantly add to the cost. We find this to be somewhat mystifying, as in the much busier and equally overburdened criminal system it is common to have bifurcated trials involving a trial on the facts and a later bifurcated sentencing hearing. This trial proves that such a process is indeed possible and reasonably practicable. We suggest that it should be considered in some civil trials. It is argued that the *Ahmed v. Stefaniu* case has limited precedential value. Morris17 explained that a review by the appellate court is limited to questions of law. The court cannot question findings of fact unless it can be demonstrated that no reasonable jury, properly instructed by the trial judge, would come to such an opinion. The bar is set very high to reach this conclusion. In other words, the Court of Appeal concedes that it did not see and hear the witnesses and concedes that it is in no position to argue with the trier of fact, who was present during the trial. Therefore, Morris concludes that the Court of Appeal does not stand for the proposition that Stefaniu was negligent, even though it was not prepared to set aside the jury's finding. It is noted that, even if this decision is not binding, it can be persuasive. Other judges can decide to follow the case and in fact, deciding according to precedence is one of the very foundations of our legal system. Cases on a similar level generally have precedential value, as it is helpful for the law to be predictable. Judges try to follow the persuasive reasoning of other judges. Jury verdicts do not produce written reasons. Therefore, there are no detailed analyses by which another jury or judge could be persuaded. Juries on the whole have only a fact-finding role and do not consider previous cases. We suggest that other plaintiffs and their counsel would be encouraged by this verdict, even if it does not bind any other court. Morris,17 in an article subtitled “Is the sky really falling?” argues that we do not know whether mental health professionals in a similar situation would face the same civil responsibility. He notes that in *Ahmed v. Stefaniu*, the Court of Appeal commented that it would not address duty of care since the matter was not raised in the appeal. Morris makes the assumption that the Court of Appeal was giving a hint that if the issue had been raised, it could have been significant. Morris cites a case in Australia18 that had some similarities. In this case, a patient discharged himself from a psychiatric facility and only six hours afterward killed his fiancée's brother. In the judgment, public policy was cited as an important consideration influencing the verdict. The trial judge found that the health authority did not breach the duty of care. The Court of Appeal in a long and complex decision addressed the nature and content of the duty of care in that case. Regarding public policy, the concept of deinstitutionalization of psychiatric services has been embraced by governments in North America and Europe for over 40 years and has resulted in a significant reduction in bed capacity for psychiatric patients in Canada.19 Expenditures on psychiatric services for inpatient facilities in Canada decreased from $250 million in the late 1980s to $104 million by 1999.19 One hundred million dollars of this savings was invested in community programs, allowing for inflation and population growth, a significant decrease in spending. Savage and McKague20 advocate that the central value to mental health legislation should be that of a least-restrictive alternative and that legislation should clearly state that institutionalization should be resorted to only as a last choice. In a recent report, Kirby and Keon21 stated that mental health service must be community based. They concluded from their hearings that many people achieve better outcomes when treated in the community. Their ninth recommendation encouraged the government to create a mental health transitional fund to accelerate the transition to a system in which delivery of mental health services is based predominantly in the community. ## Commentary on Implications of*Ahmed v. Stefaniu* The plaintiffs' experts in *Ahmed v. Stefaniu* appeared to concede that Stefaniu looked at the right parameters in making her judgment, but they believed that she gave the wrong emphasis to the factors. One conceded that the judgment was made honestly and intelligently, although this admission, which came during cross-examination, was left for the jury to assess. It is clear to every practicing mental health expert that even if judgment is exercised honestly and intelligently, we are going to get it wrong in some circumstances, simply because human behavior is very difficult to predict and the circumstances and social context are likely to change. It is the natural course of psychiatric illnesses to ebb and flow. The Court of Appeal did not address the interrupted chain of causation in which two other psychiatrists saw the patient in the seven weeks between his discharge and the killing, suggesting that the chain of causation had been broken. It seems that it behooves the Court of Appeal to analyze a point that had escaped the minds of a jury who could be swayed by advocacy and emotion. We would have expected a much clearer attempt to analyze the law by such an eminent court. We argue that the Court of Appeal should dismiss the purely emotional arguments and make a decision on the correct interpretation of the law, while considering social policy. We have grave concerns that the Court of Appeal chose to rubber stamp a decision that has serious implications. We have long learned to trust and respect this court and many in the profession will echo our sentiment of betrayal at this decision. The decision in *Ahmed v. Stefaniu* could have a very disturbing effect on physicians who assess patients and may encourage physicians to detain a patient longer than is necessary for fear of liability.11,12 Makin,11 a leading justice journalist, noted that both the psychiatric profession and psychiatric survivors' movements were united in their opinion that this decision will encourage psychiatrists to retain patients in the hospital, if there is any cause for doubt. He reports that a leading mental health lawyer and activist has already noted this trend. He concludes that this case may well swing the pendulum back toward institutionalization, especially with hard to diagnose patients who seem dangerous one day, but appear safe the next. We argue that if the Canadian government, acting through the provinces, wishes to encourage the process of deinstitutionalization and the subsequent reduction in psychiatric beds in concert with the closing of mental hospitals, then the legal system should be prepared to take this policy into consideration. It is not unusual for the legal system to shape social policy in a variety of ways.22 We conclude that the decision in *Ahmed v. Stefaniu* will have a chilling effect on psychiatrists and encourage them to practice defensive medicine. This defensive trend will further overburden a psychiatric system that has a stated intention to discharge patients to the community but is faced with a system that makes physicians liable for these decisions. Moreover, if such cases result in longer hospitalizations, the capacity of the system will be easily outstripped, and the social reform agenda of the past 40 years will be seriously threatened. This case therefore represents an aberration of the social policy judgment that the community is in most cases more therapeutic than the hospital. ## Acknowledgments We thank Ms. Linda Wolsley for her dedication and helpful comments on the article, Bill Black and Kirk Makin for their help, and Ms. Christie Hayos for her attention and aid. * American Academy of Psychiatry and the Law ## References 1. Picard EL, Robertson GB: Legal Liability of Doctors and Hospitals in Canada (ed 4). Toronto: Thomson Carswell, 2007 2. Tarasoff v. Regents of the University of California, 529 P.2d 553 (Cal. 1974) 3. Tarasoff v. Regents of the University of California, 551 P.2d 334 (Cal. 1976) 4. Resnick PJ: Duty to protect, in Advances in Forensic Psychiatry. Santa Ana, CA: CME, Inc. (undated) 5. Wenden v. Trickha, 116 A.R. 81 (QB. 1991) 6. Smith v. Jones, 1 S.C.R. 455 (Can. 1999) 7. O'Shaughnessy R, Glancy G, Bradford J: Canadian landmark case: Smith v. Jones, Supreme Court of Canada: confidentiality and privilege suffer another blow. J Am Acad Psychiatry Law 27:614–20, 1999 [FREE Full Text](http://jaapl.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6MzoiUERGIjtzOjExOiJqb3VybmFsQ29kZSI7czo1OiJqYWFwbCI7czo1OiJyZXNpZCI7czo4OiIyNy80LzYxNCI7czo0OiJhdG9tIjtzOjIwOiIvamFhcGwvMzcvMi8yNTAuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 8. Chaimowitz G, Glancy G: CPA position paper: the duty to protect. Can J Psychiatry 47:1–4, 2002 [PubMed](http://jaapl.org/lookup/external-ref?access_num=12355688&link_type=MED&atom=%2Fjaapl%2F37%2F2%2F250.atom) 9. Chaimowitz G, Glancy G, Blackburn J: Duty to warn. Can J Psychiatry 45:899–904, 2000 [PubMed](http://jaapl.org/lookup/external-ref?access_num=11190358&link_type=MED&atom=%2Fjaapl%2F37%2F2%2F250.atom) 10. Ahmed v. Stefaniu, O.J. No. 4185 (2006) 11. Makin K: Who killed Roslyn Knipe? Toronto Life. November 2007, pp 81–6 12. Willems M, Robertson D: Psychiatrist liable for actions of patient released from detention under the Mental Health Act. Faskin Martineau Health Law Bulletin January, 2007, pp 1–4 13. Snell v. Farrell, 25 S.C.R. 311 (1990) 14. Athey v. Leonate, 35 S.C.R. 458 (1996) 15. Aristorenas v. Comcare Health Services, SCC 31760 CA (2006) 16. Pengelley PD, O'Connor C: Foresight in hindsight: an insight into Ahmed vs. Stefaniu and legal evaluations of reasonable care. Available at [http://ssrn.com/abstractequals1158586](http://ssrn.com/abstractequals1158586). Accessed November 22, 2008 17. Morris JJ: Ahmed v. Stefaniu: is the sky really falling? J Ethics Ment Health 2:1–4, 2007 18. Anor v. Preslend, Hunter Area Health Service, SWCA 33 (2005) 19. Sealy P, Whitehead PC: Forty years of deinstitutionalization of psychiatric services: an empirical assessment. Can J Psychiatry 48:249–57, 2004 20. Savage H, McKague C: Mental Health Law in Canada. Toronto, ON: Butterworths, 1987 21. Kirby MJL, Keon WJ: Out of the shadows at last: transforming mental health, mental illness and addiction services in Canada. Standing Senate Committee on Social Affairs, Science and Technology, 2006 22. Regehr C, Glancy G, Bradford J: Canadian landmark case: R. v. Mills. J Am Acad Psychiatry Law 28:460–4, 2000 [PubMed](http://jaapl.org/lookup/external-ref?access_num=11196257&link_type=MED&atom=%2Fjaapl%2F37%2F2%2F250.atom)