The subjective consequences of suffering a first episode psychosis: trauma and suicide behaviour
- First Online:
- Cite this article as:
- Tarrier, N., Khan, S., Cater, J. et al. Soc Psychiat Epidemiol (2007) 42: 29. doi:10.1007/s00127-006-0127-2
- 579 Views
The subjective impact of a psychotic breakdown can be profound, potentially resulting in loss of social roles, hopes and aspirations and leading to stigmatisation, trauma and elevated suicide risk. This study aimed to assess the subjective effect and consequences of suffering a first episode of psychosis. It was hypothesised that suicide behaviour would be associated with the negative consequences of psychosis and co-morbid symptomatic-PTSD.
Patients were assessed by means of a semi-structured interview on their reactions and experience of their psychotic episode and its treatment and by means of standardised methods for psychotic (PANSS) and trauma-related (CAPS) symptoms.
A total of 35 patients suffering their first episode of psychosis were interviewed. As a result of the onset of their illness, 77% indicated they had suffered loss or disruption to their life, 60% had thwarted future aspirations, 38% had suffered violence or harassment, 53% had suffered stigma and 50% social exclusion. Totally, 80% felt they had been traumatised by their treatment and 38% were cases for symptomatic-PTSD. Symptomatic-PTSD was significantly associated with involuntary hospitalisation but not psychotic symptoms. Positive psychotic symptoms were associated with harassment, stigma and social exclusion. Suicidal ideation was reported by 40% and 31% reported attempting suicide. Suicidal behaviour was greater in those suffering symptomatic-PTSD but this was not significant, suicidal behaviour was significantly associated with the experience of trauma, but not the severity of that trauma, prior to the onset of their psychosis.
The negative consequences of a psychotic episode are significant. The potential iatrogenic effect of psychiatric care needs to be considered. Interventions need to be developed to reduce traumatisation and suicide risk.
Keywordsfirst episode psychosisschizophreniasubjective effectssuicide risktraumaposttraumatic stress disorder
There has been considerable recent interest in the way people react to the experience of a psychotic episode and in how it affects their lives and aspirations (Birchwood 2003). The experience of a psychotic breakdown can result in significant personal consequences both in the short and long term. Initially the psychotic experience will result in confusion and intense emotional and potentially traumatic reactions as a result of disturbances of thought and perception, which will be further compounded by the effects of hospitalisation, restraint and treatment. In the context of a confused state of mind the actions of the psychiatric services can provoke intense fear and helplessness. In the longer term the effects of psychosis and its potential disruption to functioning and the capacity to fulfil various social roles can result in further burden. The advent of a psychotic illness will also most certainly result in a change in the way the sufferer perceives themselves and their ability to achieve goals and in the way they are perceived by others. The societal reaction can further adversely affect the sufferer’s self-image and achievement motivation, and in some cases may result in hostile and rejecting social reactions (Angermeyer and Matschinger 2003, 2005).
Subjective feelings of social inferiority both as a direct result of psychotic experience (in attributing dominance to the voices, Gilbert et al. 2001) and due to stigma and adverse social reactions may result in demoralisation, depression and suicide risk. The latter being associated with loss, insight and fear of mental deterioration (Hawton et al. 2005) and with hopelessness, low self-esteem and social isolation (Tarrier et al. 2004); with risk being high early in the illness (Palmer et al. 2005)
The upsurge of interest in the effects of trauma in those suffering from psychosis, has been both in the traumatising effect of the psychotic experience and its treatment and in the experience of other traumatic events. Mueser et al. (2004) summarised studies investigating the prevalence of PTSD in patients suffering from severe mental illness with reported rates ranging from 14% to 43%, the lowest rate being in a study of first episodes (14%). These studies aggregate estimates for those suffering severe mental illness in different diagnostic groups. Rates of PTSD in those suffering schizophrenia tended to be lower in most studies than those in patients suffering from other disorders, for example 28% in schizophrenia compared to 54% in depression (Seedat et al. 2003) and lower still (10%) in those suffering first episodes of schizophrenia (Neria et al. 2002). Although Kilcommons and Morrison (2005) recently reported PTSD rates of 53% in a small (n = 32) convenience sample of chronic schizophrenic-spectrum disorder patients. They also found that that the severity of the trauma was associated with severity of PTSD and psychosis. Rates of PTSD in the general population are estimated at 1–14% for lifetime prevalence (APA, 1995, p 437). Thus rates in first episode schizophrenia and psychosis do not appear to be markedly different from the lifetime prevalence in the general population, although different time courses may underestimate comparative rates for psychosis. Those who suffer severe mental illnesses are more likely to be exposed to traumatic events than the general population. Mueser et al. (1998) reported 98% and Kilcommons and Morrison (2005) reported 94% of their samples having experienced at least one major traumatic event in their lifetime. There was an average lifetime exposure to 3.5 and 3.6 different events, respectively, in these two studies.
Although the experience of psychosis is undoubtedly traumatic there are a number of problems that may arise with diagnosing PTSD as a consequence of an illness, especially a mental illness, which is an internal event rather than an externally occurring, and usually an objectively verifiable traumatic event (Tedstone and Tarrier 2003). Difficulties arise in accurately describing and distinguishing aspects of the psychopathology represented by mental phenomena and avoiding double counting of symptoms and so artificially inflating estimates of co-morbidity. Furthermore controversy remains as to the reliability and validity of trauma reports in those who suffer severe disturbances of mental state (Seedat et al. 2003). The psychosis may influence and distort the patient’s recall and memory, or recall could be entangled within the delusional system (Tarrier 2005). There is also an increasing tendency to over diagnose PTSD in general (Middleton and Shaw 2000; Summerfield 2001) and the use of questionnaires can be particularly problematic (Tedstone and Tarrier 2003). Despite these difficulties, Mueser et al. (2004) reviewed the evidence for the reliability and validity of the assessment of trauma and PTSD in those suffering from severe mental illness and conclude that such assessment could be performed rigorously. In spite of the estimates of relatively high prevalence, co-morbid PTSD was rarely documented in the clinical notes of schizophrenic patients suggesting that the effects of traumatic experience are rarely investigated or recognised. Mueser et al. (1998) reported that only 3 of the 119 patients (2%) identified with PTSD in their sample had a chart record. Similarly low rates of report were cited by Kilcommons and Morrison (2005). These results indicate that the effects of trauma are rarely investigated in psychotic patients.
The risk of suicide in schizophrenia is high and is the major cause of premature death amongst schizophrenic patients (Caldwell and Gottesman 1990; Fenton et al. 1997). Most recent estimates based on meta-analysis indicate 4.9% of schizophrenic patients will commit suicide during their life, usually near illness onset (Palmer et al. 2005). Suicidal ideation and planning are important steps that lead to an attempt of self-harm that may lead to death (Kontaxakis et al. 2004) with previous unsuccessful suicide attempts increasing risk for later successful suicide (Hawton et al. 2005). Besides being a risk factor for successful suicide, suicidal ideation and behaviour are in themselves exceedingly distressing and impair functioning. In a sample of recent onset (within 3 years) schizophrenic patients, 25% reported a current desire to kill themselves and 47% had made one or more attempts (Tarrier et al. 2004). There is also evidence of elevated suicide behaviour in people suffering from PTSD (Tarrier and Gregg 2004). Thus the effects of traumatisation and co-morbid PTSD may further add to an already high risk in psychotic patients. Furthermore, co-morbid PTSD in those suffering from psychotic disorders may impair response to conventional treatment and require additional treatment approaches (see Mueser et al. 2004). Finally if the possible traumatising aspects of treatment and hospitalisation are understood then it may be possible to change practice so as to alleviate these.
It is important to know and understand the possible effects of traumatic experience in psychotic patients, especially if this is caused or exacerbated by psychiatric services or treatment. It is also important to understand how the onset of a psychotic illness will affect the subjective perception of future goals, aspirations and social position. The negative perceived effect of a psychotic illness on social acceptance and future aspirations may well influence the person’s beliefs about themselves and result in despondency, hopelessness and possibly suicidal ideation. The initial impact of a psychotic episode can best be investigated during the aftermath of the first episode.
The present study investigated the psychological effects of experiencing a first episode of psychosis and its treatment, especially hospitalisation. We were interested in investigating how people felt that their life had been changed by their experiences and especially in their perception of adverse effects such as loss, thwarted aspirations and in stigmatisation, harassment and exclusion. We investigated the traumatic effects of the psychosis and its treatment and screened for post-traumatic stress symptoms. We investigated the effects on suicidal behaviour both of trauma associated with the psychotic breakdown and also with traumas prior to the first episode. Specifically, we tested the following hypotheses: that suicidal behaviour would be associated with greater traumatisation and negative consequences of the psychosis and its treatment. These hypotheses were indicated by previous research cited earlier demonstrating elevated suicide behaviour in people suffering from PTSD (Tarrier and Gregg 2004) and in those perceiving they would suffer significant adverse or deteriorating consequences of psychosis on their lives (Caldwell and Gottesman 1990). Further, we investigated which aspects of their treatment participants found traumatic and the reasons they gave for contemplating suicide.
Patients hospitalised for their first episode of non-organic psychosis were recruited into the study. Inpatient units in Manchester were regularly contacted for potentially eligible participants who were approached by their key worker. The study was part of a larger audit of first episode psychosis in Manchester Mental Health and Social Care NHS Trust. Recruitment took place between November 2003 and April 2004. Ethical approval was granted by the local medical research ethics committee. Patients who agreed to take part in the study were then approached and assessed by the research assistant (SK, JC).
A semi-structured interview was developed to cover the following items: consequences of psychosis; reactions to treatment; traumatic reactions to psychosis; and suicidal behaviour.1 Each item was scored on being absent or present and if present then rated on severity as mild to moderate (1) or severe (2). The interviewees were asked specifically to respond to the way their psychosis had affected their lives outside of hospital, so as to avoid possible confusion with the experience of psychosis and hospitalisation. The latter being covered in a specific part of the interview.
Consequences of psychosis
Participants were asked whether, as a result of the psychosis, they had experienced significant: (1) persistent loss, change or disruption to their life (in general rather than resulting from hospitalisation); (2) reduced hopes and aspirations for the future; (3) physical harassment or violence; (4) stigma; (5) social exclusion. If present they were asked to rate the severity of the experience as mild to moderate (1) or severe (2).
Reactions to treatment
Participants were asked whether they experienced trauma as a result of: (1) being hospitalised; and, (2) the treatment they had received and to rate the severity of trauma as mild to moderate (1) or severe (2). Participants were asked what aspects had been traumatic for them.
Participants were initially screened as to whether they had experienced a significantly traumatic reaction as a result of their hospitalisation or treatment that had had a marked and persisting emotional affect. Those who gave a positive response were then assessed with the Clinician Administered PTSD Scale (CAPS) modified for use with psychotic patients (Gearon et al. 2004). This was used to provide a symptomatic diagnosis of PTSD caseness and total post-traumatic stress symptoms and intrusion, avoidance and arousal sub-scale scores. That is only symptom categories B (reliving), C (avoidance/numbing) and D (avoidance) were used.
Participants were asked: (1) whether they had experienced suicidal ideation as a result of the psychosis; and to rate the severity of the experience; and, (2) whether they had attempted suicide since the onset of their psychosis. Participants were also asked about the reasons for attempting suicide.
The Positive and Negative Syndrome Scale (PANSS) (Kay et al. 1987) was used to assess psychotic psychopathology and consisted of total, positive symptoms, negative symptoms and general psychopathology sub-scale scores. Insight was assessed by means of the Insight Scale (Birchwood et al. 1994). This is a 8-item self-report scale and higher scores indicate greater levels of insight.
The semi-structured interview took between 5 and 15 min to administer whilst the PANSS to between 30 and 45 min and the CAPS 30–60 min. Demographic and clinical information was also recorded from hospital records. Duration of untreated psychosis (DUP) was calculated from an algorithm based on information obtained from patient notes, self-report and communication with staff and relatives, where available. DSM IV descriptions were used to define the psychopathology. The timing of onset was estimated conservatively for each source (e.g., a range of 10–12 months ago as 10 months). The sources were combined using the algorithm, in practice using the longest of the estimates derived from the different sources. Hence DUP was calculated as the duration of delusions, hallucinations or psychotically disorganised behaviour (whichever was longer). This estimate was used since these symptoms were taken to have the most reliably dated onset in this population.
A total of 87 patients suffering from first episode were judged as eligible for the study, of which 35 were recruited. Thirty patients refused to consent and 14 failed to respond and 8 were lost through relocation. A further 3 failed to respond to some assessments.
The mean age of the sample was 24.9 years (sd = 6.3), 25 (71%) were male, and 20 were white Caucasian, 9 Afro-Caribbean or African, 3 Asian from the Indian sub-continent, 3 from other ethnic groups. Twenty-five (71%) were detained under the mental health act during their admission. Thirty (85%) were single, 1 married, 1 divorced and 3 co-habiting. Six (17%) were employed, 21 (60%) unemployed and 8 (23%) in full-time education at the time of onset. Mean scores on the PANSS for the total group and by gender were as follows: total PANSS: 80.66 (sd = 18.66, range 42–116), male 81.24 (sd = 42.12, range 42–116), female 79.2 (sd = 10.61, range 58–98); positive symptoms: 22.17 (sd = 6.26, range 7–35), male 22.08 (sd = 6.96, range 7–35), female 22.4 (sd = 4.33, range 16–31); negative symptoms: 18.14 (sd = 7.75, range 8–36), male 19.12 (sd = 7.74, 8–36), female 15.7 (sd 7.59, 8–34); general symptoms: 40.34 (sd = 9.97, range 23–61), male 40.04 (sd = 11.16, range 23–61), female 41.4 (sd = 6.52, range 26–49).The mean score on the Insight Scale was 7.92 (sd 4.75).2 The mean and median DUP were 15.5 and 16 weeks, respectively, (sd 11.9 weeks; range 1–52) and length of hospitalisation was 67.8 and 62 days, respectively, (sd 61.8 days; range 2–286). There were no significant gender differences on any clinical measure.
Consequences of the psychotic episode
Consequences of the experience of the onset of a first episode of psychosis
Consequence due to psychosis
Persistent loss, change or disruption
Hopes and aspirations no longer achievable
Suffered physical harassment
Experience of hospitalisation traumatic
Aspects of treatment traumatic
Prior traumatic experience
Suicidal due to psychosis
Traumatisation in response to hospitalisation and treatment
Twenty-eight (80%) reported that they felt they had been traumatised by being hospitalised and 23 (66%) indicated that this was severe. The major reason for this traumatisation was being confused/scared by being hospitalised (n = 11, 31%) followed by police ‘insensitivity’ (n = 4, 11%), fear of other patients (n = 3, 9%) and adverse staff attitudes (n = 3, 9%) and being forced to take medication (n = 3, 9%). Thirteen (37%) found aspects of their treatment traumatic and eight (23%) responded that this was severe. The major reason given was being forced to take medication against their will (n = 11, 31%). Of the participants who indicated that they had been traumatised the symptomatic criteria for PTSD, as assessed by the CAPS, was met by 12 (38%).
There was a significant association between PTSD and being detained under the mental health act (chi2 = 5.98, df = 1, P = 0.019). This was reflected in those so detained had significantly more severe avoidance symptoms (means; detained = 21.6, sd = 9.9, not detained = 12.8, sd = 10.3, t = 2.25, df = 27, P = 0.03) and a trend towards more intrusive symptoms (means; detained = 16.2, sd = 7.4, not detained = 10.2, sd = 10.7, t = 1.78, df = 27, P = 0.087) but no significant differences on arousal symptoms. There was no significant association between PTSD or CAPS scores and positive, negative or general psychotic symptoms measured by the PANSS. Those with PTSD had significantly greater reduction in their hopes and aspirations (t = 2.32, df = 24.42, P = 0.029) otherwise there were no differences in loss, change and disruption to life; and the experience of stigma or social exclusion. Those who had experience of physical harassment or violence had significantly higher scores on the total CAPS (harassed: 58.4 (17.6), not harassed 32.8 (24.7); t = −3.12, df = 26, P = 0.006), avoidance (harassed: 18.1 (8.4), not harassed 11.4 (8.8); t = −2.98, df = 26, P = 0.004), arousal (harassed: 14.7 (9.2), not harassed 7.3 (7.4); t = −2.36, df = 26, P = 0.026). With intrusions approaching significance (harassed: 18.1 (8.4), not harassed 11.4 (8.8); t = −2.00, df = 26, P = 0.057). There was no difference in CAPS scores between those who had and those who had not experienced prior trauma. DUP and duration of hospitalisation were not associated with PTSD.
Eighteen (56%) responded that they had suffered a trauma prior to their psychotic episode. These varied in their severity from incidents that would be classified as severe trauma (n = 8) to more mild life events (n = 10). Examples of the severe incidents were: rape (n = 4), and being mugged and assaulted (n = 2). Less severe events were break up of a relationship (n = 4), suffering a bereavement (n = 2) and job loss (n = 1). One stated that they were drugged by their family, which may well have been delusional. There was no association between experiencing a prior trauma and loss, change and disruption to life; reduction in hopes and aspirations; experience of physical harm or violence; the experience of stigma or social exclusion.
Fourteen (40%) responded that they had experienced suicidal ideation as a result of their psychotic illness, 9 (26%) responded that this was severe and 11 (31%) indicated that they had attempted suicide as a result of their illness. Of the reasons for suicide attempts 9 were for reasons of depression, hopelessness or frustration; 1 was in response to auditory hallucinations and 1 was ‘to be with’ a significant other who had died.
There was a trend towards an association between suicidal ideation and being detained under the mental health act (chi2 = 3.23, df = 1, P = 0.073). Fifty-eight percent of those who met PTSD caseness were suicidal compared to 35% who did not meet PTSD caseness, but this association was not significant. There were no significant associations between suicidal ideation and loss, change and disruption to life; reduction in hopes and aspirations; experience of physical harm or violence; the experience of stigma or social exclusion. There was a significant association between suicidal ideation and the experience of prior trauma (chi2 = 4.88, df = 1, P = 0.036), although this was not related to the severity of the trauma. Suicidal ideation and attempts were not associated with positive, negative, general symptoms or depression on the PANSS or DUP and duration of hospitalisation.
We investigated the effects of suffering a psychotic breakdown on a sample of first episode patients. The sample was relatively small but their characteristics were not dissimilar to those recruited to another recent onset schizophrenia project carried out in the north of England (the Socrates trial; Lewis et al. 2002; Tarrier et al. 2006) which suggests that this sample was not unrepresentative. There were reports of marked negative consequences to the patients of suffering a psychotic breakdown, both in terms of persistent loss (77%) or significant impediment to achieving their aspirations (60%) and also in how they felt themselves perceived by others. Over a third reported physical harassment of some kind and these were more likely to experience significantly higher post traumatic stress symptoms. A further 50% or more felt stigmatised or socially excluded.
Positive psychotic symptoms were significantly higher in those with reports of harassment, stigma and social exclusion but without objective corroboration it is not possible to say whether these aversive consequences were accurate recall or resulting from the state of their mind due to positive symptoms. However, as these groups did not significantly differ on the PANSS delusion sub-scale this may indicate this perception is not solely a product of mental state. DUP and time in hospital were not significantly associated with subjective consequences, which again suggest that severity of illness was not the main explanation.
A very definite finding was that many patients find the experience of the episode and their treatment as deeply distressing. A clear majority reported being traumatised by being hospitalised and a sizeable minority by the treatment they received. Thirty-eight percent satisfied symptom criteria for symptomatic-PTSD, which did not appear to be related to the severity of psychotic symptoms. Levels of suicidal ideation and reported suicide attempts were high.
We found little evidence to support our first hypothesis that suicidal behaviour would be associated with the negative and subjective consequences of psychosis. Although a sizeable proportion of the sample reported that they had experienced persistent loss, change or disruption to their lives; felt that their hopes and aspirations were no longer achievable; had suffered physical harassment or violence; and suffered stigma or social exclusion these factors were not significantly associated with suicidal ideation or attempts. This is perhaps surprising because pessimism about the future would be expected to be associated with hopelessness and suicide risk. Possibly these are factors that become more important after recovery from the episode rather than during it. Suicide behaviour was higher in those identified as symptomatic-PTSD cases, although this association was not significant. There was however, a significant association between suicidal ideation and the experience of a trauma prior to the onset of psychosis. This was not related to the severity of the prior trauma although the low subject numbers may have meant that there was not the statistical power to find a difference. Possibly the onset of psychotic illness further added to stress and despondency resulting from previous traumas which had sensitised the individual to subsequent stress. It is possible that the lack of an association between suicidal behaviour and other factors is because suicidal behaviour may increase with time since Tarrier et al. (2004) reported that suicidal ideation and behaviour increased with time over the 3 years from the onset of the first episode.
The onset of a psychotic illness results in a number of significant changes in the person’s perception and expectation about their future. As with any life event or illness these beliefs can add to the burden of the event by creating a negative frame of mind which discourages positive and goal directed behaviour. The appraisal of the psychotic experience and its consequences has been shown to be associated with subsequent emotional responses (Birchwood et al. 1993). Many patients experiencing their first episode of psychosis experienced or perceived adverse social reactions even though there has been a very recent onset of their illness. Clearly there are potentially dramatic changes in how the person perceives themselves as a result of this episode. It is possible that these biases in cognition relating to self and social knowledge may sensitise the individual to interpret future anomalous or stressful experiences in an increasingly psychotic manner. The nature of these appraisals may confer a psychological vulnerability, which increases the probability of subsequent psychotic episodes. Morrison (2001) has linked traumatic experience to later psychosis through mediation of such a cognitive bias.
It is important to understand patient’s beliefs and expectations and their consequences as these may be effectively targeted in cognitive-behaviour therapy, especially early in the illness. Generic CBT may successfully reduce psychotic symptoms (Pilling et al. 2002; Tarrier and Wykes 2004) but not significantly reduce suicide behaviour (Tarrier et al. 2006). Thus a more detailed understanding of belief formation that may lead to suicide behaviour and despondency about the future may be important in designing specialised CBT for the reduction of suicide risk and a positive approach to recovery after the first episode.
The study has a number of limitations. The sample size was modest and there was a high refusal rate in those eligible to participate which may also indicate a bias in those who agreed to participate although that only three participants failed to complete the study means that the drop-out rate was relatively low. The high refusal rate may have been because patients were too distressed by their experience to want to talk about it or because they were disengaged from, or dissatisfied with, the mental health services. Standardised methods were used to assess clinical symptoms but a semi-structured interview was used to assess participant’s psychological reaction to their psychosis. In spite of a standardised assessment interview specifically modified for psychotic patients to detect PTSD it was apparent that it is not easy to identify PTSD using the full criteria. This was especially true relating to the stressor, thus we are of the opinion that caution should be used in diagnosing co-morbid conditions. There may be an argument for identifying a specific sub-type of post traumatic stress that accompanies psychosis so as to acknowledge this diagnostic difficulty and to differentiate traumatic stress as a result of psychosis and its treatment from traumatic reactions to a clear external event. Moreover, the scoring of the semi-structured interview itself may have imposed a restrictive range of responses and thus have reduced sensitivity.
Future research should prioritise the further investigation of subjective experience of psychosis and its consequences and in attempting to reduce the traumatising and potentially iatrogenic effects of psychiatric intervention; the development of psychological treatments to ameliorate the effects of trauma and general demoralising effects of a emergence of a psychotic illness and to reduce suicide risk; and in understanding the course and long term effects of these factors.