Feigned psychosisrevisited-a 20 year follow up of 10 patients

more commonly before the courts, and diagnosis required identification of a specific motive such as attempts to avoid prosecution, conscription or punishment. Ganser (1898) described three cases of an hysterical twilight state in prisoners and concluded that these were not the result of malingering but true illness. Jung (1903) stressed what he considered to be the strong relationship between criminality, "malingering", a term which he left undefined, and "simulation", actions

Early accounts of feigned insanity included observations relating to its detection by a lack of the particular odour believed to attend the truly insane (Hill, 1814).So called "pretenders to madness" (Beck, 1829) were said to be found more commonly before the courts, and diagnosis required identification of a specific motive such as attempts to avoid prosecution, conscription or punishment.Ganser (1898) described three cases of an hysterical twilight state in prisoners and concluded that these were not the result of malingering but true illness.Jung (1903) stressed what he considered to be the strong relationship between criminality, "malingering", a term which he left undefined, and "simulation", actions intended to deliberately conceal inner healthi ness.He found only 11 malingerers among 8340 admissions to hospital, but no fewer than nine of these patients had been investigated for or convicted of a crime.Slater (1961) described how the diagnosis of hysteria may be indicative of the nature of the relationship between a particular doctor and patient at a specific time and in certain circumstances.He also found that there was frequently significant and serious underlying physical or major psychiatric disorder in such cases, sometimes only diagnosed accu rately years after the original presentation despite the presence of clear physical findings or signs throughout the intervening period (Slater, 1965).This is in keeping with the suggestion that feigned psychosis may have its roots in genuine psychia tric disorder and that where the diagnosis has been made actual illness may emerge later (Hay, 1983).
Shakespeare described how Edgar in King Lear feigned insanity and took on the guise of "Poor Mad Tom".Paradoxically, since his madness was simulated, the detail of the account of Poor Tom's life and condition have been cited as evidence that Shakespeare must have been familiar with chronic schizophrenia and that the illness, contrary to what had been suggested previously, was indeed known in the sixteenth century (Bark, 1985).
The simulation of mental illness had been used as means of escape from prisoner of war camps in the First and Second World Wars (Reid, 1952;Jones, 1955) and featured in fictional writing (Schneck. 1970).Samuel Fuller's film Shock Corridor in which a journalist attempts to unravel an unsolved murder in an asylum by feigning insanity only to be overtaken by true mental illness, was considered to be in such bad taste that its screening was originally banned for 7 years by the British Board of Censors.More recently media attention has focused upon the issue of serious offenders attempting to pervert the course of justice by the simulation of mental disorder.Szasz (1987) has explored the imitation of mental illness as well as the supposed deception involved in any apparent remedy and cites Swift's reference in Gulliver's Travels to imaginary diseases and imaginary cures.
Modern diagnostic classifications include facti tious disorder with feigning of psychological symptoms characterised by apparently obscure internal motivation and malingering associated with more obvious external stresses or incentives (World Health Organization, 1992;American Psychiatric Association, 1994).There remains doubt nevertheless about the diagnostic legiti macy of simulated mental illness (Jonas & Pope, 1985;Rogers et al 1989).Pope et al (1982) emphasised the features often present in cases of factitious psychosis, in particular the almost universally poor outcome, but argued that their findings in relation to a 4 to 7 year follow up period suggested that the diagnosis remained valid.In contrast Hay (1983) identified a group of six patients who fulfilled his study criteria for simulation, those discharged over a 10 year period with a diagnosis of feigned psychosis or remembered by the responsible doctor as having feigned psychotic illness, all but one of whom later developed some form of genuine disorder.
In the present paper we review the original diagnosis and describe the outcome for 10 patients who presented with what were appar ently simulated psychotic symptoms more than 20 years ago.
The study Ritson & Forrest (1970) described 12 patients admitted to a psychiatric hospital who were apparently simulating symptoms of psychosis.In three cases a diagnosis of schizophrenia had already been made, although for each of these patients the presentation at the time of the episode described was considered to be characterised by feigned, rather than genuine, symptoms.Among the other nine patients there was said to be no prior history to suggest psychosis although a consistent pattern of various degrees of disturbance of personality was described in each.From the details contained in the original paper we were able to identify 10 of these 12 patients and examine their case notes.Clinical diagnosis and demographic data were recorded from the time of the previously reported episode and also any diagnoses made subsequently.The computerised instrument OP-CRIT (McGuffln et al 1991), which generates diagnoses according to a number of classificatory systems from case note, clinical or other defined sources of information, was applied for each index admission and then all subsequently recorded data.
Table 1 shows the original diagnosis for each patient at the time of the index admission, the current clinical situation, the OPCRIT ICD-10 and DSM-III-R diagnoses relating to case note information from the time of the index admission and from analysis of subsequent data.CÃ-infcaÃ-diagnoses and outcome Of the three patients said to have had a prior schizophrenic illness but feigned symptoms, two are now in long term in-patient hospital care with a clinical diagnosis of schizophrenia.The other patient has returned to her country of birth but OPCRIT suggested the presence of an affective illness at the time of the index presentation rather than schizophrenia.All of the remaining seven patients have subsequently attracted a clinical diagnosis of either schizophrenia (n=3), a major affective illness (n=3), or in one case, both of these at different times in the past.The period which elapsed between the index admission with appar ently feigned symptoms and the ultimate diag nosis of a manic depressive or schizophrenic illness ranged from 7 months to in excess of 20 years.
Of the three patients who went on to be newly diagnosed as suffering from a genuine schizo phrenic illness each has had numerous admis sions to hospital during the intervening period, but always with the same diagnosis.One is now living independently, taking long-term oral antipsychotic medication and has regular contact with a community psychiatric nurse; another lives at home, attends the out-patient department and receives intramuscular long acting medica tion; and the third is in supported accommoda tion, also being treated with a depot neuroleptic preparation.
Two of the three patients who subsequently attracted a diagnosis of affective disorder have had episodes of both hypomania and depression.Both had a family history strongly suggestive of serious mental illness.One of these two had no contact with psychiatric services in the interven ing 20 year period following the index admission then presented with her first manic episode.The third patient in this group developed a depressive illness two years after being admitted with what were thought to be simulated symptoms.The remaining patient attracted diagnoses of hypomania and depression, but on other occasions schizophrenia, during the time following his first contact and before his suicide.

Standardised diagnoses
Only two patients did not fulfil criteria for OPCRIT DSM-II1-R diagnosis at the Urne of their index admission, but all met ICD-10 criteria for some form of disorder at that point (Table 1).There were no follow up data available in one case where the patient initially met criteria for mania with psychosis, but the remaining nine patients all fulfilled criteria for major mental illness when all subsequent episodes were included.Seven of these had schizophrenia according to both classiflcatory systems but two had other diag noses, in one case bipolar disorder by both ICD-10 and DSM-III-R criteria, and in the other, non-organic psychosis and bipolar disorder respectively.

Comment
This study is unusual in that it allowed a follow up period of at least 20 years for all those concerned.The finding that in most cases sufficient signs and symptoms were recorded at the time of the initial presentation to meet operational criteria for a major disorder by OPCRIT diagnosis is of particular interest.
Despite the obvious limitations of such a retrospective study based only upon case note data there was apparent consistency between the ultimate clinical diagnosis and that generated by the objective instrument in each instance.In considering the concept of diagnosis it has been held that such consistency and persistence over time, in keeping with the natural history of the disorder, substantiates the potential validity of the diagnostic entity.This would seem not to be the case for "feigned psychosis".One of the patients described here was lost to follow up.Two more did not meet DSM-III-R criteria initially, but were ascribed diagnoses of mild and moderate depression respectively at that time according to ICD-10.The remaining seven had evidence of a diagnosis of major affective disorder or schizophrenia stable over time and a natural history and response to treatment in keeping with this.
The fact that all these patients developed a major psychiatric illness over the course of the follow up period is similar to the findings of Hay (1983), although the emergence of bipolar dis order in two cases is an outcome not previously reported.It seems likely that so called feigned psychotic symptoms may indeed represent tran sient or more enduring phenomena, genuinely experienced and reported, which develop at some later date into a more clearly defined and recognisable pattern.
On the basis of the present investigation it is not possible to determine how the patients described in the study might have differed at the time of their original presentation from other psychotic patients.It is unclear why, when case notes made at the time were sufficient to fulfill standardised criteria for a psychotic disorder in the majority of cases, symptoms were attributed to feigning.There was evidence of some specific motivating factor for the patient, in the form of social or domestic upheaval, in two of the cases included in the present study but only one faced a criminal charge at the time of index admission.
All of those who had not previously attracted a clinical diagnosis of schizophrenia were said to have had personality problems or a personality disorder.This may be strongly associated with the notion that an individual is in control and entirely responsible for their actions and might diminish the significance of symptoms otherwise recognised as genuine (Lewis&Appleby, 1988).It is also noteworthy that although on occasion the content of psychotic symptoms may be under standable in psychodynamic terms, the diagnos tic importance of their presence remains.
The long-term outcome findings presented from this study lend further weight to the view that the utmost caution must be exercised in regard to the suggestion that psychotic symptoms might be simulated, or at least that they may have little significance in terms of the future development of the clinical picture, especially at the time of first presentation to psychiatric services.The subse quent onset of a more clearly recognisable major mental illness remains likely.The evidence for the diagnosis of feigned psychosis may be no more objective than the absence of the odour described by Hill (1814) and lead to the same potential for clinical error as the diagnosis of hysteria (Slater, 1965).In our opinion the term should be abandoned.It might be better to make a detailed description of the reported phenomena in terms of "atypical symptoms" and to keep an open mind about their significance and any future diagnosis regardless of what treatment might be deemed appropriate at the time (Scott, 1965).

Table 1 .
Original and subsequent diagnoses for each patient, and their current clinical situations