Abstract
Keywords
Psychotic disorders with acute onset, dramatic symptomatology, short duration and a generally good prognosis have long been recognized as clinical entities, although both diagnostic labels and underlying concepts have varied [1–3]. In 1992 these disorders were included in the 10th edition of the International classification of mental and behavioural disorders (ICD-10), under the label of acute and transient psychotic disorders (ATPD) [4]. This category is defined as a psychotic disorder of acute onset (within 2 weeks) and of limited duration (not longer than 3 months). For a diagnosis of ATPD, major affective disorders, organic disorders and schizophrenia have to be excluded. Differentiated psychopathological features are used to define several subgroups, including ‘acute polymorphic psychotic disorder’, with or without symptoms of schizophrenia (F23.0 and F23.1), ‘schizophrenia-like psychotic disorder’ (F23.2) and ‘acute and transient psychotic disorder, predominantly delusional’ (F23.3). Preceding stress as a trigger for the episode is not necessary for diagnosis, but if present may be coded additionally. Thus, the definition of the general category of ATPD is rather broad, but provides criteria for several subtypes. This structure reflects the nature of ATPD as a consensus category integrating concepts of different regional schools and traditions [3].
The oldest, and most influential, source for ICD-10 ATPD is the concept of ‘bouffée délirante’. In French psychiatry the diagnosis of bouffée délirante has a history of more than 100 years [2, 5, 6]. In the 1880s, Valentin Magnan (1835–1916) and his pupil Maurice Legrain (1860–1939) described for the first time a psychopathological condition named ‘syndromes épisodiques de les dégénérés’ or ‘bouffée délirante des dégénérés’ [7, 8]. Magnan, the ‘grey eminence’ of French psychiatry [9] and one of the most important French nosologists, was devoted to the study and propagation of the idea of ‘degeneration’, introduced by Bénédict-Augustin Morel in 1857 [10]. Magnan extensively described the psychiatric disorders of hereditary degenerates. He put particular stress on the occurrence of specific psychotic illnesses of the degenerate. Often these disorders take the form of psychotic disorders of acute onset and rapid remission, termed ‘délire d'emblée’ (sudden madness) or ‘bouffées subites d'idées délirantes’ [7].
The clinical picture as summarized by Pichot [2, 6] includes (i) a sudden onset, ‘like a bolt from the blue’; (ii) delusions characterized by numerous, diverse and protean delusional themes without recognizable structure and cohesiveness, sometimes accompanied by hallucinations; (iii) clouding of consciousness associated with emotional instability; and (iv) a rapid return to the premorbid level of functioning. ‘By definition, transitory delusional states are short-lived.’ Relapses may occur, but are separated by symptom-free intervals.
At the end of Magnan's life, and with the diminishing influence of the theory of degeneration early in the 20th century, the concept of bouffée délirante was largely abandoned. In issues of classification, French psychiatry was for some time influenced by the new doctrine of Emil Kraepelin [5]. It was Henri Ey (1900–1977) who renewed interest in the concept of bouffée délirante, which he refined and contrasted to a more narrowly defined concept of schizophrenia [11]. The diagnosis conforms with the desire of many French psychiatrists to put more emphasis on course than on symptomatology and to separate the acute psychoses from schizophrenia [12]. Bouffée délirante has a firm place in French psychiatry as documented by its inclusion in the national classification system INSERM [13] and the formulation of operational criteria by Pull et al. in 1983 [6, 12, 14]. Case reports, theoretical articles and clinical studies on this concept continue to be published mainly in France and francophone countries [15–22].
Although bouffée délirante played an important role in the creation of ICD-10 ATPD, little is known about the concordance of ICD-10 ATPD and bouffée délirante. We performed a prospective and longitudinal study on a group of 42 patients who fulfilled the ICD-10 criteria of ATPD, the Halle Study of Brief and Acute Psychoses (HASBAP) [23]. In order to establish the relationship of ATPD and bouffée délirante, we conducted an analysis comparing patients with ATPD who concurrently fulfilled the criteria of bouffée délirante with those who did not. In this paper we set out to answer two questions: (i) What concordance exists between the ICD-10 diagnosis of ATPD and a diagnosis of bouffée délirante?; and (ii), do patients with bouffée délirante differ in prognosis from other patients with ATPD?
Method
Sample
The sample described in this study is composed of 42 inpatients with ATPD treated at the Department of Psychiatry and Psychotherapy of Martin Luther University Halle-Wittenberg between 1993 and 1997. The details of recruitment and data acquisition have been described elsewhere [24, 25]. Briefly, patients with a clinical discharge diagnosis of ATPD were considered for inclusion in the study. All diagnoses were reviewed on the basis of a checklist incorporating ICD-10 research criteria. Consensus was reached on each patient. Only subjects in whom a diagnosis of ATPD was confirmed were included in the study. Additionally, a diagnosis of bouffée délirante was made according to the criteria of Pull et al. (1983) [14].
Age at onset: approximately between 20 and 40 years
Onset: acute, without any prior psychiatric history (other than identical episodes)
No chronicity: active phases fade away completely in several weeks or months possibly recurring in the same form: the patient remains devoid of abnormality in the interval
Characteristic symptoms (all of the following):
–Delusions and/or hallucinations of any type
–Depersonalization/derealization and/or confusion
–Depression and/or elation
–Symptoms vary from day-to-day, even from hour-to-hour
Not due to any organic mental disorder, alcoholism, or drug abuse.
Assessing the index episode
Data from the clinical records and patients' answers to specified questions at the follow-up interview were entered on a form designed to assess several features of the index episode. These features included sociobiographic data, previous episodes, symptomatology and the time course of the index episode. Associated acute stress was rated as present in a restrictive fashion according to ICD-10 criteria (i.e. only when the first psychotic symptoms occurred within about 2 weeks of one or more events that would be regarded as stressful to most people in similar circumstances).
Follow-up
We undertook follow-up examinations of all living and consenting patients. Follow-up investigations took place 8.2 ± 8.3 years (mean ± SD) after the first episode (range = 0.8–27.3 years), or 2.2 ± 1.3 years after the index episode (range = 0.6–5.1 years). At the time of follow-up, one patient had died, three refused consent and the remaining 38 subjects (90.5% of the original sample) were personally interviewed. Subjects who dropped-out did not differ significantly from subjects with completed follow-up on a wide range of sociodemographic and clinical variables. Follow-up measures included a semistructured interview for the evaluation of sociobiographic features that had been used in earlier studies. Past episodes of illness with in- or outpatient treatment were recorded. As an episode, we counted the occurrence of a major affective syndrome or of psychotic symptoms that led either to hospitalization or to outpatient treatment, including psychiatric medication, and disrupted the patient's daily activities. Episodes fulfilling ICD-10 criteria at follow-up as determined by the World Health Organization Schedules for Clinical Assessment in Neuropsychiatry (WHO-SCAN) [26] were also included. All patients were investigated and interviewed by the authors themselves (four psychiatrists and a clinical psychologist). As extensive and specific data on illness and treatment history were obtained during the interview, investigators could not be blind to the diagnosis.
Outcome measures
The instruments used were the SCAN [26], the WHO Psychological Impairments Rating Schedule (PIRS) [27] and the WHO Disability Assessment Schedule (DAS) [28], all in their German translations, as well as a semistructured interview for the evaluation of sociobiographic features already used in earlier studies. Best-estimate ICD-10 diagnoses were assigned to all episodes during follow-up using all available information, including patients' reports and hospital records. The level of general functioning was assessed by means of the widely used Global Assessment Scale (GAS) [29].
Present state ICD-10 diagnoses at follow-up were assessed with the SCAN [26]. All interviewers were extensively trained on the use of this instrument. Training on the WHO instrument SCAN was carried out by WHO-approved trainers, and the first 10 interviews were supervised.
Reliability
Reliability estimates for the central outcome variables were obtained in our group. For this purpose, 15 interviews were double-coded by two raters (interviewer–observer method). Kappa values for categorical items exceeded 0.80 for all items. For quantitative outcome scales, intraclass correlation coefficients were excellent, ranging from 0.86 (GAS) to 0.94 (DAS global score).
Data analysis
Statistical analyses were performed using the Statistical Package for Social Sciences (SPSS), version 9.0. For comparison of continuous data we used two-tailed t-tests. Contingency tables of categorical data were analysed with χ2 tests or Fisher's exact test as appropriate. The level of significance was set at 0.05.
All patients provided written informed consent. The study protocol was approved by the local ethics committee.
Results
Acute and transient psychotic disorders
We found 42 cases of ATPD (4.1%) among 1036 patients treated for non-organic psychotic or major affective disorders (F2 or F3 of ICD-10) during the study period. The patients with ATPD formed 8.5% of all patients treated for non-organic psychotic disorders (F2 of ICD-10). Subgrouping according to ICD-10 categories showed that polymorphic features, schizophrenic symptoms, or a combination of both were present in the majority of cases. There was a marked female preponderance in ATPD, with a female to male ratio of 3.7: 1. The sex ratio in ATPD differed significantly from that in patients treated for schizophrenia (ICD-10 F20) during the inclusion period (female to male ratio in schizophrenia 0.7: 1; significance of difference: χ2 = 21.58, df = 1, p < 0.001).
Concordance of ATPD with bouffée délirante
When diagnosed according to the criteria of Pull et al. [14], a minority of cases (n = 12) were classified as bouffée délirante. This was 28.6% of all patients with ATPD, 1.2% of all patients treated for nonorganic psychotic or major affective disorders (F2 or F3 of ICD-10) during the study period, or 2.4% of all patients treated for non-organic psychotic disorders (F2 of ICD-10). Nearly all patients with bouffée délirante belonged to the polymorphic subtype of ATPD (10/12, 83.3%). However, many patients with a polymorphic subtype of ATPD did not meet the Pull et al. criteria of bouffée délirante (18/28, 64.3%). Thus, concordance between the two diagnoses is far from complete. Bouffée délirante as operationally diagnosed can best be described as a subgroup of ATPD, but much less voluminous.
Comparison between two groups of acute and transient psychoses: bouffée délirante and non-bouffée délirante
We divided the sample into two groups: those patients who fulfilled the Pull et al. [14] criteria of bouffée délirante and those who did not. None of the parameters of the index episode differentiated significantly between bouffée délirante and non-bouffée délirante. Associated acute stress tended to be more frequent in bouffée délirante (Table 1).
Baseline data and characteristics of index episodes of 12 cases of bouffée délirante compared to 30 cases of acute and transient psychotic disorder not fulfilling the criteria of bouffée délirante (non-bouffée délirante)
Data concerning the course of the disorder up to the index episode are also given in Table 1. Some patients had previous episodes, the highest number being 16. Retrospective assessment of earlier episodes with operational criteria was hampered by variable, poor and sometimes absent documentation. A preliminary classification of these episodes according to ICD-10 showed ATPD episodes to be most frequent (47%), followed by affective (24.2%) and schizoaffective (15.2%) episodes. The only significant difference between patients with bouffée délirante and those without concerned age: in bouffée délirante age at onset and age at index episode were lower. This difference is compatible with the operational definition of bouffée délirante, which requires onset by up to approximately 40 years of age. In contrast, in ATPD, onset was not restricted to any specific age group but could occur as late as 70 years. No differences were found in course parameters between bouffée délirante and non-bouffée délirante patients.
Follow-up
We carried out personal follow-up interviews at a mean of 8.2 years after the first episode and 2.2 years after the index episode. The attrition rate was low, at less than 10% of the original sample. In the period between index episode and follow-up, 29 of 38 patients (76.3%) experienced a relapse. The number of relapses ranged from 0 to 4, with a mean of 1.18 and a median of 1. The number of relapses did not differ significantly between the bouffée délirante sample and the non-bouffée délirante sample (Table 2). Diagnostic status at follow-up was assessed with SCAN interviews (present state). Schedules for Clinical Assessment in Neuropsychiatry diagnoses at follow-up are given in Table 2. The diagnoses shown in this table are highly operational, but as presentstate diagnoses they are of cross-sectional nature. Within these limits, they can be regarded as an indicator of diagnostic stability. Two patients met the diagnostic criteria of schizophrenia (ICD-10 F20); both belonged to the non-bouffée délirante group. Two additional patients of each group suffered from a depressive episode (F32). One patient in the bouffée délirante group was in an episode of ATPD at follow-up. The majority of patients in each group did not fulfil the criteria of a psychotic or major affective disorder at follow-up and thus could be considered to be in syndromal remission.
Episodes during follow-up period and SCAN diagnoses at follow-up and Global Assessment Score a follow-up in 12 cases of bouffée délirante and in 26 cases of acute and transient psychotic disorder not fulfilling the criteria of bouffée délirante (non-bouffée délirante)
We also investigated psychosocial variables as markers of psychosocial functioning and treatment status. Consistent with the considerable relapse rate, most patients were on psychotropic (mainly neuroleptic) medication at follow-up (bouffée délirante: n = 7, 58.3%; non-bouffée délirante n = 21, 80.8%; difference not significant). Interestingly, patients with bouffée délirante were significantly more frequently employed at follow-up (9/12, 75.0%) than those without a diagnosis of bouffée délirante (9/26, 34.6%; p = 0.02). The difference diminished only slightly when patients older than 60 years were excluded from the analysis. Patients with bouffée délirante were also more often in a stable heterosexual partnership at follow-up (10/12, 83.3%) than those without a diagnosis of bouffée délirante (13/26, 50.0%), but the differences were not statistically significant.
Using standardized instruments, we evaluated outcome in three additional domains: social disability, psychological impairment and global functioning. Global Assessment Scale scores as a measure of global functioning disability are given in Table 2. There were no significant differences between bouffée délirante and non-bouffée délirante patients, although the bouffée délirante group scored slightly better. In general, the outcome was favourable; 11 of the patients with bouffée délirante (91.7%) and 20 of the patients without bouffée délirante (80.8%) showed a GAS score greater than 70 at follow-up. The pattern of DAS scores (social disability) and PIRS scores (psychological impairment) paralleled the results obtained with the GAS (data not shown).
Discussion
When the operational criteria of Pull et al. [14] were applied to the present cohort of 42 patients with ATPD, only 28.6% of the sample were concurrently diagnosed as bouffée délirante. Thus, the operational definition of bouffée délirante is much narrower than that of ATPD. In consequence, patients with bouffée délirante in this sample are rather rare and represent only 1.2% of all patients with non-organic psychotic or non-organic major affective disorders, or 2.4% of all patients with non-organic psychotic disorders. This finding is in stark contrast to the continuing popularity of the diagnosis bouffée délirante among French clinicians; it is assigned to approximately 16% of psychotic patients [18, 21, 30]. However, the diagnosis seems less frequent in research settings [18]. Singer et al. [15], in their follow-up investigation of 74 patients initially diagnosed with bouffée délirante, noted that only a narrow subgroup of patients (n = 19) fulfilled the strict criteria of Pichot [31]. In addition, when taking into account the follow-up data, these authors concluded that only 0.47% of psychiatric admissions were ‘true’ bouffée délirantes. The criteria suggested by Pull et al. [14] require, among others, acute onset, a specified polymorphic syndrome, full remission and absence of different psychotic episodes, all of these obligatory. Although these criteria were derived from the empirical study of the diagnostic practice of French psychiatrists, it seems that, in everyday use, the criteria are followed less strictly than an operational procedure would require. Some reports on bouffée délirante have included psychotic states induced by drugs or sleep deprivation, these were not included in the present investigation [20]. Drug induced psychoses may show strong phenotypic similarities to bouffée délirante.
The female preponderance in the present sample supports the separation of bouffée délirante from schizophrenia, which is equally represented in both genders or even slightly more frequent in males [32, 33]. As a result of the age restriction included in the definition of Pull et al. [14], age at onset in the present sample with bouffée délirante was lower than in the non-bouffée délirante group. For the combined group of ATPD, however, age at onset was higher than in unselected samples of patients with schizophrenia [34, 35]. Thus, contrary to the assumptions of Pull et al. [14], young age does not seem particularly characteristic for acute brief psychoses, and its inclusion in the definition of bouffée délirante is not supported by the present data.
During the 2-year follow-up, attrition was rather low (less than 10%). In this period, the vast majority of both bouffée délirante and non-bouffée délirante patients experienced a relapse. Due to the naturalistic character of the study, we did not control for intervening factors that might have influenced the course of the illness during follow-up, such as medication and other forms of treatment. In fact, the majority of patients received some form of maintenance treatment, mainly antipsychotics. The high relapse rate is in accordance with the concept of bouffée délirante [6] and other brief and acute psychoses [24]. Some studies, however, reported similar relapse rates for cohorts of patients with schizophrenia and schizoaffective disorders (e.g. the 81.9% relapse rate found by Robinson et al. [36]). A more detailed analysis of the present data showed that the majority of episodes during the follow-up period could be classified as ATPD, but affective and schizoaffective episodes occurred as well [25]. Two patients showed diagnostic change to schizophrenia. Despite frequent relapse, most patients in the present sample outcome were excellent in terms of social functioning and symptomatology. The mean GAS score at follow-up (79–83) was somewhat higher than the GAS of 70 reported by Jørgensen et al. [37] in their 1-year follow-up of a cohort of ATPD patients and considerably higher than scores found in follow-up studies of patients with schizophrenia (e.g. 44.7 in [38]). Full remission was attained by 80–87% of the present sample when a moderately strict criterion (GAS > 70) was used, while comparable remission rates in follow-up studies on schizophrenic samples range from 21% to 57% [39]. Some patients, however, showed mild to moderate degrees of impairment. Taking all findings together, the generally favourable outcome in the present sample confirms the good prognosis implied by a diagnosis of ATPD [40, 41].
As the operational criteria for bouffée délirante result in a narrow selection of patients, the question arises as to whether these patients are also selected for a particularly favourable prognosis. In the present data, there is indeed some indication that this is the case. Patients with bouffée délirante were significantly more often employed at follow-up, they were more often in a stable heterosexual partnership, and they showed a slightly more favourable outcome on most standardized outcome scales. The small number of subjects, however, meant that these differences did not attain statistical significance. The problem of low statistical power is not peculiar to the present study, but rather results from the low prevalence of bouffée délirante, as discussed above. Since the present samples were selected from 1036 inpatients, to achieve a higher statistical power, a study would need to screen a very high number of patients. Yet, a replication of these findings in a larger sample is desirable.
Various considerations, however, argue for the validity of the present findings. First, there is a strong converging trend in the outcome measures pointing in the direction of a more favourable prognosis of bouffée délirante. Second, the present findings are compatible with results from a separate analysis of the HASBAP. This analysis showed that a concurrent diagnosis of cycloid psychosis (which also includes a polymorphic symptom picture and sudden onset as defining criteria) also carried the implication of a more favourable prognosis [42]. Third, the present findings are concordant with the study of Singer et al. [15] mentioned above. These authors compared the course in 19 patients diagnosed with bouffée délirante according to Pichot [31] with that in 28 patients diagnosed with bouffée délirante on clinical grounds. It was noted that evolution towards schizophrenia was less common in narrowly diagnosed bouffée délirante.
Conclusions
Bouffée délirante shows substantial overlap with ATPD but has a considerably lower frequency than ATPD when diagnosed with operational criteria. There are indications that a diagnosis of bouffée délirante carries a somewhat better prognosis than ATPD in general. The differences, however, are slight. The low frequency of operationally diagnosed bouffée délirante indicates that the criteria of Pull et al. [14] may be too narrow.
Footnotes
Acknowledgements
This study was supported by a grant from the German Research Council (DFG MA 115/12–1).
