Skip to main content
PMC home page
The BMJ logoLink to The BMJ
. 2001 May 26;322(7297):1301.

Post-traumatic stress disorder

Disorder takes away human dignity and character

Arieh Y Shalev 1
PMCID: PMC1120389  PMID: 11403051

Editor—Encouraged by Summerfield's revelation about post-traumatic stress disorder,1 I imagined myself going to my clinic the next day and, at last, telling my patients who have post-traumatic stress disorder that their disorder is but social invention. I also thought that I would apologise, admitting that I was wrong in choosing to diagnose their problem and thereby medicalise their condition instead of seeing it as normal human suffering. Given that suffering is normal, as Summerfield says, I was also prepared to encourage my patients to be happy with having survived adversity and never again mention the word victim. It is a matter of dignity. Better be normal and suffer than have a mental disorder treated.

My daydreaming continued, and I saw myself meeting an anorectic patient, for whom I care very much, and telling her that given the social roots of her disease in ideals of feminine thinness, she should simply start eating. To my very sick schizophrenic patient, overwhelmed by demonic exacerbation, I was to show that schizophrenia is but a scientific delusion.2 So far for social constructivism.

If anything, the birth of post-traumatic stress disorder exemplifies how good it is that despite orthodoxy and haughtiness the medical profession is sometimes forced to listen to people's pain. Not that post-traumatic stress disorder is built in stone. But neither are depression, psychosis, or delirium. Meanings change with time, and I hope that this will continue. What is, however, fascinating in post-traumatic stress disorder is that, despite its tentative beginnings, this diagnosis has generated more replicable biological findings than many traditional disorders.3 Moreover, the development of post-traumatic stress disorder in traumatised people offers a major opportunity to study the ways in which mental events transform the central nervous system.4 The marriage between post-traumatic stress disorder and the neurosciences seems more productive than the disorder's acceptance in some circles.

I wish to protest, once again, against the reluctance to identify a mental disorder in those who suffer, just because this might become a psychiatric diagnosis. I thought that those days were over; that human dignity is not lost when one has a mental disorder.

Doctors should encourage their patients to disclose distress and seek help. In their daily practice they should and can discern normal sorrow from major depression, doubt from obsessive rumination, idiosyncrasy from schizophrenia, and transient responses to extreme events from post-traumatic stress disorder. They have nothing to gain from claims that the pervasive and interminable personal disaster that is post-traumatic stress disorder is not a disorder.

References

  • 1.Summerfield D. The invention of post-traumatic stress disorder and the social usefulness of a psychiatric category. BMJ. 2001;322:95–98. doi: 10.1136/bmj.322.7278.95. . (13 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Boyle M. Schizophrenia: a scientific delusion. London: Routledge; 1990. [Google Scholar]
  • 3.Pitman RK, Shalev AY, Orr SP. Post-traumatic stress disorder: emotion, conditioning, and memory. In: Gazzaniga MS, editor. The cognitive neurosciences. 2nd ed. Cambridge, MA: MIT Press; 2000. pp. 1133–1148. [Google Scholar]
  • 4.Shalev AY, Pitman RK, Orr SP, Peri T, Brandes D. Auditory startle in trauma survivors with post-traumatic stress disorder: a prospective study. Am J Psychiatry. 2000;157:255–261. doi: 10.1176/appi.ajp.157.2.255. [DOI] [PubMed] [Google Scholar]
BMJ. 2001 May 26;322(7297):1301.

Doctors should relieve suffering, not debate its existence

Juliet Cohen 1

Editor—In his views on the sociological origins and implications of post-traumatic stress disorder Summerfield seems to overlook the clinical basis for this complaint.1-1 There are real patients who complain of symptoms best described by the diagnostic criteria of the Diagnostic and Statistical Manual of Mental Disorders, fourth edition, and that is why these are used. There is a gradation of traumatic experience, from minor road traffic accident to horrendous atrocity, just as there is a gradation from sadness to full blown depression. That does not make either less of a clinical entity. Dismissing the suffering of a patient with post-traumatic stress disorder as a sociological problem seems like telling a depressed patient to pull himself or herself together.

Summerfield quotes the American Journal of Psychiatry when saying that if anyone liked a psychiatric diagnosis they were given it would be post-traumatic stress disorder. Has he ever really listened to a patient describe the hell of his or her nightmares and flashbacks? Has he looked at his or her pallor, red rimmed eyes, bitten fingernails and thought that this was merely a construct of media hype and compensation neurosis? Perhaps he has been lucky enough never to have an accident or witness any horror in his medical training that had the power to linger in the memory and reappear in dreams? A diagnosis lacking specificity is hardly unique in medicine. Arthritis is a condition with an enormous range of symptoms and severity, but the label itself is not without use both to doctors and patients.

Summerfield also comments on the dependence of post-traumatic stress disorder on external events to define its onset and the recency of this idea as pathological in the context of the long history of man's traumatic memories. Most orthopaedic diagnoses are entirely dependent on external traumatic events for their causation, but we do not find surgeons agonising over whether to nail broken pieces of bone back together just because once they were left to heal as best they could.

Summerfield cites a recent survey in Freetown, Sierra Leone. He comments that the finding of a 99% incidence of post-traumatic stress disorder is clinically meaningless, presumably simply because the incidence is so high. But is it not possible the incidence is high because terrible events have occurred and the people are suffering? It may be possible to define war as a sociological construct, but its effects on human beings are as real as the people who experience them. It is up to doctors to try to relieve this suffering, not debate its existence.

References

  • 1-1.Summerfield D. The invention of post-traumatic stress disorder and the social usefulness of a psychiatric category. BMJ. 2001;322:95–98. doi: 10.1136/bmj.322.7278.95. . (13 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 2001 May 26;322(7297):1301.

Logic is flawed

Simon J Ellis 1

Editor—Summerfield in his article on post-traumatic stress disorder starts with flawed logic and ends in denial.2-1 By starting from an assumption that a psychiatric diagnosis has an objective existence independent of the observer he sets psychiatry in a world of its own. All other diagnoses are observer dependent. All reality is observer dependent. Migraine exists as an illness and may be related to changes of serotonin. It is irrelevant whether neolithic people suffered with what we would call migraine. The concept of migraine is useful in clinical practice. Its existence is dependent on its utility. Homosexuality was once considered a disease. In today's society such a view is laughable. The concept of homosexuality as a disease has lost its utility and so no longer exists.

Summerfield believes that post-traumatic stress disorder is a recent social construct, despite citing evidence that something similar (shell shock) was recognised during the first world war. He doubts that neolithic people had post-traumatic stress disorder and therefore denies its existence. I doubt that neolithic people had much in the way of squamous cell lung cancer, but I do not doubt the utility of such a diagnosis today.

Summerfield thinks that the idea of a traumatic memory being a pathological entity is a recent construct. This concept goes back at least as far as Freud.2-2 It seems not unreasonable that memories of bad events might produce psychological harm. To believe otherwise would mean having to reject the notion that sexual abuse in childhood can result in psychological damage and social dysfunction in adulthood. If you believe in relativity, then, although there may be societal norms of what constitutes a traumatic event, the crucial issue is how an event was perceived by the individual.

Summerfield believes that post-traumatic stress disorder confabulates normality and pathology and devalues “true” illness. The criteria for post-traumatic stress disorder given in the Diagnostic and Statistical Manual of Mental Disorders, fourth edition, specify that the disturbance causes clinically significant distress and impairment in social, occupational, and other important areas of functioning.2-3 This is not a description of normality.

Summerfield contributes to the myth that there is an army of professionals supporting unjust claims for damages. The reality is that litigation for personal injury is difficult, slow, and the financial compensation is generally much less than the financial loss. I am yet to meet a claimant who would rather have the money they finally receive than the injury they sustained. He criticises “sympathetic” psychiatrists for assisting claimants. Far better some sympathetic psychiatrists assisting the courts in determining as accurately as possible the true disabilities of claimants than unsympathetic psychiatrists deny suffering and disability on the grounds that they do not think Neanderthal man had similar problems. If a school of psychiatry wishes to distance itself from such patients so be it, but other caring professionals will not.

References

  • 2-1.Summerfield D. The invention of post-traumatic stress disorder and the social usefulness of a psychiatric category. BMJ. 2001;322:95–98. doi: 10.1136/bmj.322.7278.95. . (13 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2-2.Breuer J, Freud S. Studies on hysteria. New York: Basic Books; 2000. . (Originally published 1895.) [Google Scholar]
  • 2-3.American Psychiatric Association. Diagnostic and statistical manual of mental disorders DSM-IV-TR (text revision). 4th ed. Washington, DC: APA; 2000. [Google Scholar]
BMJ. 2001 May 26;322(7297):1301.

Social usefulness of any diagnosis needs consideration

Andrea Litva 1

Editor—I do not believe that Summerfield's argument is dangerous.3-1 He is simply applying the social constructionist model to post-traumatic stress disorder and this ideology is no more threatening or destructive than the biomedical model.

The social construction of illness and disease is well recognised in the literature on medical sociology, and Summerfield's article adds to an existing debate. The socially constructed nature of (almost?) all illness or disease or disorder is well recognised among medical sociologists and psychologists. Illness and disease cannot exist or emerge separate from society. What is perhaps downplayed in the article and seems to have aroused much angst among respondents is a recognition of the importance of diagnosis—any diagnosis—for patients who have a collection of symptoms.

The sociologist Talcott Parsons first recognised the importance of going to the doctor and being diagnosed in his description of the sick role. Regardless of the sociopolitical and medical contexts from which illness or diseases or disorders emerge (although fascinating), in Western society when people's conditions are diagnosed by clinicians the people are generally no longer regarded as personally responsible for causing their illness behaviour. Arguably this removes them from being regarded as deviants or at least morally questionable people and places them firmly into the legitimating hands of medicine. For those with post-traumatic stress disorder, a diagnosis can provide them with much needed legitimacy, allowing them to retain their place within society without diminishing what they are experiencing. It may well be the medicalisation of everyday life, but as long as the biomedical approach dominates the way Western society approaches their problems, it is a very necessary evil. Perhaps we need to start thinking beyond clinical outcomes and start thinking what the real role of medicine is in Western society.

References

  • 3-1.Summerfield D. The invention of post-traumatic stress disorder and the social usefulness of a psychiatric category. BMJ. 2001;322:95–98. doi: 10.1136/bmj.322.7278.95. . (13 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 2001 May 26;322(7297):1301.

Accountable methods of validation are needed

Grant J Devilly 1

Editor—Summerfield's article is an example of good intention, poor method, and bad outcome.4-1 Post-traumatic stress disorder is not solely a legacy of the American war in Vietnam. For example, see Herodotus' Aristodemus (“the trembler”) in The Histories, or Lady Percy's speech in Shakespeare's Henry IV, Part 1 (scene 3).

The kernel of truth to this section of Summerfield's argument is, however, that the Vietnam war led to a ready number of returned servicemen displaying and reporting similar problems. Because of the similarity of symptoms shared with victims of sexual assault, however, this was not called war neurosis.

Research requires criteria that ensure comparison of similar presentations. This description (post-traumatic stress disorder), which we shall call a diagnosis, helps make research studies comparable across settings. Debriefing people after a disaster—a process of iatrogenically presenting, priming, and prepping medicalised information—now seems to be a very poor response.4-2 Summerfield warned of this as far back as 1995.4-3

However, people presenting with a cluster of signs (symptoms?) and describing this with a term does not equate to the medicalisation of a problem. Negating it can be a disenfranchisement of the patient. If this is a socially constructed phenomenon, why do we get similar presentation rates, symptoms, and success rates after treatment around the world? When we do not arrive at similar treatment success rates this alerts us to a possible problem with either the sample or the intervention.4-4

Summerfield also seems to be suggesting an argument against the current form of trauma tourism being practised by so many debriefing companies and humanitarian assistance programmes. This is laudable, but different to suggesting that all diagnoses are a Western concept and therefore harmful.

Inhabitants of war torn Sierra Leone are not an example of people with post-traumatic stress disorder. The trauma is current, and the symptoms are, therefore, adaptive. Delineating between true cases and malingering is again an issue of which all therapists and forensic assessors are aware. Yet to dismiss all cases of post-traumatic stress disorder, as Summerfield suggests, as examples of people with a flaccid upper lip is risible. The need for accountable methods of validation is, however, well taken. I question whether Summerfield believes the presented argument himself, but rather is playing Devil's advocate for the sake of debate.

References

  • 4-1.Summerfield D. The invention of post-traumatic stress disorder and the social usefulness of a psychiatric category. BMJ. 2001;322:95–98. doi: 10.1136/bmj.322.7278.95. . (13 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4-2.Kenardy J. The current status of psychological debriefing: It may do more harm than good. BMJ. 2000;321:1032–1033. doi: 10.1136/bmj.321.7268.1032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4-3.Summerfield D. Debriefing after psychological trauma. Inappropriate exporting of western culture may cause additional harm. BMJ. 1995;19:509. doi: 10.1136/bmj.311.7003.509a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4-4.Devilly GJ, Foa EB. The investigation of exposure and cognitive therapy: comments on Tarrier et al (1999). J Consult Clin Psychol (in press). [DOI] [PubMed]
BMJ. 2001 May 26;322(7297):1301.

Baby should not be thrown out with bath water

Tom Clark 1

Editor—Summerfield discussed the validity of the diagnosis of post-traumatic stress disorder.5-1 He overstated his case, however, and diluted the impact of his argument by confusing two fundamentally distinct issues. The concept of illness and the validity of medical diagnoses do not necessarily impinge on a consideration of the potential misuse of such diagnoses. Post-traumatic stress disorder is overdiagnosed, having expanded beyond the medical sphere, to those of psychologists, nurses, other health professionals, lawyers, and the media. In modern society, the influence of prevailing individualistic and self centred values and the need for a universal term for suffering that may justify the actions of “victims,” litigants, and their advocates may well have contributed to this overexpansion of the concept.

This does not, however, invalidate a diagnosis when used appropriately. AIDS is an obvious example (of many) that denies Summerfield's assertion that an illness must have existed in neolithic times for it to be real. The influence of sociopolitical factors in the genesis of psychological symptoms that may form a diagnostic category similarly does not deny the diagnosis. Psychological, social, and other extrinsic factors are of importance in the pathogenesis of all illnesses, physical as well as psychiatric. Such factors may give rise to illness just as legitimately as genetic aberrations, infection, diet, and lifestyle. Consequently illnesses may indeed come and go through time and across cultures. Summerfield's questions, “Where were they before?” and, “Where did they go?” remain valid, but they do not imply a lack of objective reality.

The reliance of psychiatry on a syndromal classification of disease introduces subjectivity to diagnosis and a vulnerability to misuse and misunderstanding. This is not a problem unique to psychiatry. The problems described by Summerfield may also operate with regard to diagnoses of whiplash injuries after road traffic accidents, post-concussional syndrome, back injuries, chronic fatigue syndrome, and many others. It is not a coincidence that such complaints are commonly the subject of petition and litigation.

The baby (post-traumatic stress disorder) should not be thrown out with the bath water (spurious eponymous diagnoses). Rather, psychiatry should strive to classify psychological reactions to stress more accurately, pursue the holy grail of biological markers to inform and aid appropriate and objective diagnosis, and embrace the task of educating the public and the media in the correct significance and interpretation of current diagnostic labels.

References

  • 5-1.Summerfield D. The invention of post-traumatic stress disorder and the social usefulness of a psychiatric category. BMJ. 2001;322:95–98. doi: 10.1136/bmj.322.7278.95. . (13 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 2001 May 26;322(7297):1301.

Questions about current status of psychiatric classification systems arise

Casimiro Cabrera-Abreu 1

Editor—Summerfield's paper on post-traumatic stress disorder was long overdue.6-1 As a psychiatrist working in Canada and having been trained in Great Britain, I have been overwhelmed by the arbitrariness and the superficiality of the construct “post-traumatic stress disorder,” which I encounter almost everyday tagged to some of my patients in an irresponsible fashion.

It is from two Canadian scholars, a philosopher and a historian of psychiatry, that we can get some further insight into Summerfield's incisive critique of this condition. The philosopher, Ian Hacking, in a recent book borrows from another philosopher, John Searle, two concepts that can be applied to Summerfield's analysis.6-2 In discussing the social construction of mental illness, Hacking mentions the fact that young women with anorexia nervosa die from their condition. In this sense their condition is epistemologically objective but ontologically subjective. Post-traumatic stress disorder is also epistemologically objective—people suffer—but it displays ontological subjectivity—it is the result of an elaborate historical process mired in the intricate tapestry of the construction of the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised, and its sequels.

The second Canadian scholar is Edward Shorter, who, in his book From Paralysis to Fatigue, impinges upon the subtle and interactive process between doctors, patients, and cultural mores.6-3 Spinal irritation, reflex theory, gynaecological surgery to cure nervous mental illness, and hysterical fits were the result of this ongoing process. According to Shorter, when the doctor's idea of a “legitimate” disease changes, the patient's idea changes as well. When the doctors shifted their paradigm from reflex neurosis emphasising motor hysteria to the central nervous paradigm of sensory symptoms, the patients shifted accordingly.

I believe that Summerfield's paper elevates the tone of the debate around the unwarranted medicalisation of human suffering rather than confusing it. It also poses questions concerning the current status of psychiatric classification systems.

References

  • 6-1.Summerfield D. The invention of post-traumatic stress disorder and the social usefulness of a psychiatric category. BMJ. 2001;322:95–98. doi: 10.1136/bmj.322.7278.95. . (13 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6-2.Hacking I. The social construction of what? Cambridge, MA: Harvard University Press; 1999. [Google Scholar]
  • 6-3.Shorter E. From paralysis to fatigue. A history of psychosomatic illness in the modern era. New York: Free Press; 1992. [Google Scholar]
BMJ. 2001 May 26;322(7297):1301.

Diagnostic label is misleading

Annie Hargrave 1

Editor—I am in agreement with practically everything Summerfield says in his article on post-traumatic stress disorder.7-1 I work with emergency, aid, and development workers returning from overseas and regularly meet the pain and anguish associated with war, disaster, violence, and suffering.

I would like to see a more rigorous definition of the word trauma. I use it in my mind to mean a piercing of defences and resources, both conscious and unconscious, that is experienced, for a time at least, as overwhelming. It is not applicable to the frustration of missing the bus or the disappointment of failing an exam, for example.

I think of what I call ordinary post-traumatic stress. This is the normal distress and dislocation of shock and suffering, often, in my context, separated from the event itself because the individual has split it off (shelved it if you like) in order to do their job. People sometimes need help with this but the aim is to reconnect them with their own mechanisms, both internal and social, so that they can grieve, learn, and get on with living. I do not regard this as pathological at all. It is wonderful to see people recovering over just two or three appointments.

The psychosocial consequences of traumatic incidents may be difficult for people to come to terms with. It is terrible to “lose” your spouse to brain damage or be unable to work again because of injuries sustained. But this is not post-traumatic stress disorder as I understand it. Post-traumatic stress disorder, to me, is much rarer. I would use it to describe a condition of entrenched collapse of defences and resources over time. Patients from the aid and development world may present one year, five years, after a traumatic event and describe a monumental struggle to overcome difficulties which are now overcoming them. The traumatic situation or event broke through the psychological defences, liberating underlying, unresolved conflicts and terrors such that the process of grieving and healing cannot take place. In my view the logical outcome of “true” post-traumatic stress disorder, if it runs its course, is death.

I agree with Summerfield that the diagnostic label is misleading. The trauma may have triggered the illness, but it is the patient's relation to the trauma that is at the heart of the matter.

References

  • 7-1.Summerfield D. The invention of post-traumatic stress disorder and the social usefulness of a psychiatric category. BMJ. 2001;322:95–98. doi: 10.1136/bmj.322.7278.95. . (13 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 2001 May 26;322(7297):1301.

Author's reply

Derek Summerfield 1

Editor—Shalev, Cohen, and Ellis all pitch “suffering” as a form of psychopathology. This distortion may reflect the advantages that accrue to medically attested outcomes, including legitimated sick roles, as Litva and Devilly remind us. But is their objection also aesthetic: is the medicalisation of life now so natural that accounts of suffering which do not deploy the language of trauma seem to be playing down what people have gone through, and are thus distasteful? Surely it would be objectionable to victims of Hillsborough or Omagh to discover that people acknowledged their experiences by attaching a mental disorder to them. What is at issue in my paper is not the “reality” of human distress, but the fidelity of a particular psychiatric category.

Ellis should re-read my paper. The assumption that a psychiatric category has an existence independent of the observer is not mine. Psychiatric orthodoxy has always assumed that there was a set of eternal facts (the “baby”) to be teased out from their context (the “bath water”). Diseases are discovered, having always existed. Devilly provides an example when he sees post-traumatic stress disorder in the works of Herodotus and Shakespeare. A Vietnam war veteran in the United States has just published a book averring that Alexander the Great had post-traumatic stress disorder.8-1 Perhaps the most radical demonstration of the medicalisation of the past to serve the purposes of the present is the transgenerational trauma movement. Trauma is seen as an entity transmissible to people still unborn when the events took place. (This was once called Original Sin). This affords the child or grandchild of a Holocaust survivor an extra way of having personal unhappiness explained, and a victim identity legitimated.8-2

Of course other psychiatric categories also have histories: for example, that of “depression” reveals the gradual incorporation of the cultural vocabulary of guilt, energy, fatigue, and stress.8-3 It is interesting that British general practitioners have begun to put “stress,” in daily use as a folk idiom, on sick certificates instead of “depression” or “anxiety.” But post-traumatic stress disorder deserves sociological attention because it is being attached so sweepingly to experiences that till now have been seen as merely part of life. So where did the epidemic of yesterday's unrecognised, untreated cases go to? Shalev is wrong when he implies that a distinct psychobiology for post-traumatic stress disorder has now been identified. Post-traumatic stress disorder is being globalised. Why, as a professor of psychiatry, does he not concede that Western psychiatry is merely one among many ethnopsychiatries, and that post-traumatic stress disorder is a syndrome bound to Western culture?

Clark is right; a diagnostic category is not necessarily to blame if it is misused. But post-traumatic stress disorder is especially amenable to misuse because so many of its criterial features are non-specific and subjective. My own clinical experience of it was honed on over 800 assessments of refugee survivors of human rights abuses. Criteria for post-traumatic stress disorder cannot distinguish between torture and a bicycle accident, exclude a pre-existing psychiatric disorder, nor the impact of current social stressors. Its conceptual basis in supposedly immutable, pathogenic “traumatic” memory is simplistic: memory is interpretative and therefore malleable. Veterans from the Gulf war were given a standard set of questions about their combat experiences one month, and again at two years, after the war had ended. Informants generally reported significantly more traumatic exposure the second time.8-4 This shift may well be pertinent to the construction of Gulf war syndrome, which still seeks the disease status accorded post-traumatic stress disorder.

As Ellis notes, homosexuality was once classified as a psychiatric disorder, a reminder that social attitudes are liable to be recast as freestanding medicopsychological facts. A wise psychiatry is one with an element of self doubt, acknowledging the limitations of a pathology bound biopsychomedical paradigm, rooted in a mechanistic view of man, in capturing the complexity and ambiguity of human experience. As a category post-traumatic stress disorder can support some weight, and I am saying we should debate how much this is, but it cannot support the tower block that has been erected on it.

References

  • 8-1.Tritle L. From Melos to My Lai. New York: Routledge; 2000. [Google Scholar]
  • 8-2.Summerfield D. The psychological legacy of war and atrocity: the question of long-term and transgenerational effects and the need for a broad view. J Nervous and Mental Disease. 1996;184:375–377. doi: 10.1097/00005053-199606000-00008. [DOI] [PubMed] [Google Scholar]
  • 8-3.Jadhav S. Cultural origins of Western depression. Int J SocPsychiatry. 1996;42:269–286. doi: 10.1177/002076409604200403. [DOI] [PubMed] [Google Scholar]
  • 8-4.Southwick S, Morgan C, Nicolaou A, Charney D. Consistency of memory for combat-related traumatic events in veterans of Operation Desert Storm. Am J Psychiatry. 1997;154:173–177. doi: 10.1176/ajp.154.2.173. [DOI] [PubMed] [Google Scholar]

Articles from BMJ : British Medical Journal are provided here courtesy of BMJ Publishing Group

RESOURCES