Part 1: Introduction Chapter 1: Context and Issues This study sets out to look at a group of individuals who have recently experienced their first episode of psychosis, and their caregivers. The individuals with psychosis are referred to throughout this study as ‘patients’ to avoid any confusion. Chapter outlines the important issues and the debates surrounding them. Also included is an outline of the study and how the thesis is organised. 1.1 The patients The patients in this study had just experienced their first episode of psychosis. Psychosis is a brain disorder; it causes gross distortion of an individual’s mental faculties to the extent that it prevents them living a normal life. Simple tasks such as clear thinking, decision making, relating to others and controlling emotions become challenging (Chan, 1995). Studies have shown that the majority of first cases of psychosis represent schizophrenia or other schizophrenia type disorders (Edwards, Maude, McGorry, Harrigan & Cocks, 1998). However, at the first episode it is not possible to make a definitive diagnosis, the priority is to treat the psychosis. See Table 1.1 for incidence details. Table 1.1 Worldwide incidence rates of Schizophrenia and First-Episode Psychosis Incidence rate per 100,000 First-episode psychosis 15-20 cases Schizophrenia 12-15 cases (Source: McGlashan, 1998) This first-episode period of psychosis has become the focus of much research because it has been shown that delay in treatment of psychosis may have far reaching effects. First, there is greater risk to patients of life threatening behaviours, aggression or suicide (Loebel et al., 1992). Second, delays prolong the distress for families and patients (Haas, Keshavan, & Sweeney, 1994). Third, there is evidence (e.g. McGlashan & Fenton, 1993; McGlashan & Johanessen, 1996) that the initial phase of the disease may be when neurological damage occurs. For this reason, this phase has been termed the ‘critical period’ because this is where the opportunity presents itself to act early and avoid the impairments caused by subsequent psychotic episodes (Birchwood, 2000; Linszen & Birchwood, 2000). What clearly emerges from the research is that the longer the illness is untreated, the poorer the outcome for the patient (e.g. Loebel et al., 1992). From the above it can be seen that the first episode is a crucial time for intervention. For this reason, research has focused on what may be triggering psychotic episodes in order to prevent them. 1.2 The caregivers All patients in this study live at home and all caregivers are family members. It has long been acknowledged that the appearance of a mental illness in a family is a disaster for the whole emotional and functional life of that family (Terkelsen, 1987). Over time the burdens of looking after the ill relative can be enormous. Burdens can be emotional such as fear, worry, anger, sadness and grief, or practical such as disruption of family life and financial strain (e.g. Greenberg, Won Kim, & Greenley, 1997; Lefley, 1997; Solomon & Draine 1995; Veltro, Magliano, Lobrace, Morosini & Maj, 1994). However for those patients and caregivers who are experiencing their first episode of psychosis the burdens can be quite different. Caregivers have to deal with the shock and trauma of discovering that their relative has a serious mental illness. They have to go through the distressing process of having their relative diagnosed and treatment started, often against the relative’s will. Diagnosis itself may take some time so there is the added stress of uncertainty. Families may believe they are somehow responsible for their relative contracting this disease. Resulting feelings of guilt can lead them to become over-involved in the patient’s life, anticipating their every need and cushioning them from stress and from the world (Chan, 1995). This may result in the patient becoming overly dependent on the caregiver and unable to be independent. Another common reaction the family may have is to become overly critical of their ill relative (Chan, 1995). Some of the symptoms of psychosis are not obviously disease related such as poor hygiene, withdrawal, etc and patients will not be performing routine tasks or actions in the same way they did before they became ill. These behaviours can provoke criticism from an already stressed family member. In addition, caregivers may have been supporting and caring for a sick family member for some considerable time with no outside or professional help. Singapore is a multi-cultural society comprising Chinese (77%), Malay (14%), Indian (7%) and other (1.7%) (Statistics Singapore, 2003). As the majority of the population in Singapore is of Chinese background, this study focused on this group of caregivers, allowing for a more detailed investigation of the cultural issues facing this particular group. 1.3 Predictors of outcome in psychosis: issues and debates As mentioned previously it is believed that brain deterioration may be occurring with each psychotic episode. One of the goals in the treatment of firstepisode psychosis therefore, is to understand what is triggering or causing the psychotic episodes in order to prevent them, and the subsequent damage. For this reason, research has focused on factors which may be potential predictors of outcome. Outcome is either (a) onset of an initial episode, or (b) relapse after having a psychotic episode. Outcome (i.e. return of symptoms) can vary greatly. Some patients will experience only one psychotic episode and return to normal functioning. Some will have several episodes over a number of years with no impairment, whilst others will continue to experience episodes with no return to normality. Malla, Norman and Voruganti (1999) suggest that 50% of patients with schizophrenia will continue to experience symptoms and require long term support. Many factors have been suggested as predictors of outcome in first-episode psychosis such as gender, with males having a poorer outcome (Beiser, Fleming, Iacono, & Lin, 1988); depression (Shepherd, Watt, Falloon & Nigel, 1989) and early negative symptoms (Scottish Schizophrenia Research Group, 1988). The research regarding these and other factors such as early age of onset is very mixed. There are other predictive factors, however, which stand out as appearing to be more important. Stressful life events in the period before an episode have received much attention as possible predictors. Bebbington et al. (1993) found that when compared with a group from the local general population, those with psychosis had a significant excess of life events in the three months preceding the onset of psychosis. However, more recent research has focused on two other potential predictors: a negative family atmosphere and the length of untreated illness prior to meeting with a psychiatrist. These two factors will be the focus of the present study. 1.3.1 Negative family atmosphere A negative family atmosphere has been the subject of research in this area for over 40 years. This interest arose from work done by the Medical Research Council’s Social Psychiatry Unit in London during the 1950s and 1960s, when, with the introduction of more effective forms of drug treatment, people with schizophrenia were returned to the community. It was noted that some fared well in the community, whilst others did not; those who returned to their homes relapsed (experienced an episode) more than those who went into other accommodation (Brown, Monck, Carstairs & Wing, 1962). It seemed that something was occurring within the family atmosphere to bring about relapse. Further study identified criticism and emotional over-involvement (EOI) in these families as being the key factors associated with relapse in the patient (Leff & Vaughn, 1985). The name given to the family atmosphere by these researchers was ‘Expressed Emotion’ (EE; Brown, Birley & Wing, 1972) which can be defined as “a global index of particular emotions, attitudes and behaviours expressed by relatives about a family member diagnosed with schizophrenia” (Jenkins & Karno 1992, p.9). Brown and colleagues suggested that it was the added stress of the parent’s (or spouse’s) criticism that was the trigger that pushed the already vulnerable patient over the threshold to experience a psychotic episode. Since that time, the predictive ability of EE in chronic schizophrenia has been demonstrated many times and in many countries. There are a number of issues concerning EE that need to be addressed. The first is that whilst EE has been shown to be a robust predictor of relapse in chronic cases (discussed fully later in Chapter 4) the predictive ability of EE in first-episode cases is not so clear. Research so far has proved ambiguous. A second issue concerns whether EE is an emergent or existing concept. Does high EE develop in families as they go through subsequent relapses with the patient, or is it present either before or at onset of the illness (Birchwood & Smith, 1987)? First episode studies are ideal for tackling this type of question. Evidence for the existence of EE in early cases would shed light on whether EE results from chronicity of illness or is a trigger in the illness process. A third question relates to what factors are underlying EE. When families are faced with the same stressor (a mentally ill relative) what causes one family to become high EE (i.e. very critical and hostile) whilst another family remains low EE. This will be discussed separately in section 1.4. 1.3.2 Period of Untreated Illness Research into the predictive ability of EE in first episode cases may be inconclusive because it is confounded with another significant predictor of outcome in these cases, duration of untreated illness (DUI). Often there is a delay between patients experiencing their first psychotic episode and seeking treatment. This period of untreated illness has been found to be a significant predictor of outcome in first-episode psychosis. Clearly it is important to determine whether EE or DUI is the better predictor of outcome for these patients. One of the important features of the above factors is that they are malleable (unlike sex or age) and therefore ideal candidates for intervention. 1.3.3 Social Support Having outlined potential predictors of outcome, it is necessary to look at factors which may be protecting, or buffering, the patient from the stress of the negative family atmosphere. It is well established that social support plays a large role in moderating the effects of stress on physical health (B. Sarason, Sarason, & Pierce, 1990) and mental health (Perrucci & Targ, 1982) particularly depression (Harris, 1992). In a review of social support in schizophrenia, Jackson and Edwards (1992) summarise the literature in terms of four questions that they see as most important. These are: 1) reduced network size of individuals with psychosis; 2) whether the reduced network is a cause or consequence of the illness; 3) whether the patient’s personality before the illness affects the social network and 4) what this impaired network means for the prognosis of the patient. It is this last question that is the focus of this study. Will the patient’s social support network buffer the effects of the stressful family environment and allow the patient to have a better outcome. 1.4 Predictors of EE: issues and debates Whilst knowing that EE is a predictor of outcome is important, it is as important to try and understand what factors underlie EE. By knowing what factors contribute to high EE (i.e. a negative family environment) interventions can be targeted at those factors. Similarly, by knowing what attitudes or other factors contribute to low EE, interventions can be aimed to encourage those attitudes in high EE caregivers. As yet researchers are still unsure about the exact mechanisms underlying EE. However, there is agreement that what is underlying EE is a process of interaction between patient and caregiver; a complex combination of different variables such as caregiver personality, coping style and patient symptomatology or length of illness (Hooley & Gotlib, 2000). This interaction process has been clearly demonstrated. Not only high and low EE families have quite different communication styles, with caregivers high in EE presenting a less tolerant and more critical style than those low in EE (Hooley & Hiller, 2000), but the patients in these families also react differently to the caregiver. Patients from high EE families were more likely to show odd and disruptive behaviours during interactions than low EE patients (Rosenfarb, Goldstein, Mintz & Nuechterlein, 1995). So, in the case of mental illness, the patient’s difficult behaviours may be both a cause and consequence of stress. The behaviours are stressful for the caregiver, the stress makes the caregiver tired and irritable and unable to deal with the patient, who then becomes more difficult (Aldwyn, 1994). This suggests that those who deal more effectively with the stress will be less irritable and that the cycle just described will not be completed. A number of caregiver factors have been suggested as being involved in this transactional process between patient and caregiver. For instance, several differences have been found in personality and temperament between high and low EE relatives. Hooley (1998) found highly critical relatives to have a more internal locus of control compared with low critical relatives who had a more external locus of control. Comparing high and low EE relatives, Hooley and Hiller (2000) found high EE relatives to be less flexible, less tolerant and lower in empathy. Guilt proneness has also been studied in a group of relatives (Bentsen et al., 1998) with those high in EE being more guilt-prone. Some personality factors, such as flexibility, act in a protective capacity preventing the caregiver from becoming high EE. Warmth is another factor which has been found to offset critical comments in some countries, e.g. India (Leff et al., 1987), Italy (Bertrando et al., 1992) and Egypt (Okasha et al., 1994). An optimistic disposition is regarded as a protective factor in this study. Others factors concern the relative’s attitudes and behaviours. Those that have emerged as being most salient are the relative’s appraisal of, and subsequent coping with, the situation and the patient’s behaviours, and the caregiver’s control attributions regarding the illness. These factors in the caregivers, along with optimistic disposition, are the focus of this study. Whilst the relationship of these variables with EE has been demonstrated in many Western populations, their relevance for a Singaporean population has yet to be demonstrated. In addition, they have not been studied in conjunction with the personality variable of the caregiver optimistic disposition. A further point is that these factors are interrelated however for clarity they are discussed separately. These factors are now briefly outlined here: 1.4.1 Appraisal The concept of burden of care has been well researched in the area of chronic schizophrenia (e.g. Bibou-Nakou, Dikaiou & Bairactaris, 1997; Schene, van Wijgaarden & Koeter, 1998) and it has generally been found that living with a mentally ill relative poses a huge burden on the family, both subjective and objective. Caregiver burden is seen as an external demand or potential threat that has been appraised by the caregiver as a stressor (Scazfuca & Kuipers, 1996). There is strong agreement that it is the caregiver’s appraisal of the situation rather than the situation itself that is crucial (Scazfuca & Kuipers 1996; Pickett, Cook, Cohler, & Solomon, 1997). For this reason, later studies have tended to look at the caregiver’s appraisal of the burden rather than an objective measure of the burden itself. Therefore, measures of burden are regarded as measures of appraisal, indicators of the way the caregiver is appraising the situation. A relationship has been found between appraisal and EE such that those who appraise the situation more negatively (i.e. perceive greater burden) are more likely to be high EE (Scazfuca & Kuipers, 1996). 1.4.2 Coping It has been found that relatives of people with schizophrenia tend to use a broad range of coping strategies and that coping strategies were influenced by appraisal (Birchwood & Cochrane, 1990; Scazfuca & Kuipers, 1999). However, as high EE has been associated with more avoidant coping (e.g. Kuipers & Raune, 2000) it could be that worry triggered by negative appraisals may inhibit the adoption of adaptive coping strategies (Birchwood & Smith, 1987). Differences on how the relatives cope may explain the low/high EE distinction. 1.4.3 Control Attributions When caregivers observe their family member displaying illness related behaviours, they generally make a decision regarding whether the patient, or the illness, is to blame, i.e. an internal or external attribution. These attributions are important, because they are seen as predicting cognitions and subsequent 10 1992, Khoo & Bishop, 1996; Segerstrom, Taylor, Kemeny & Fahey, 1998). If optimism is not affecting outcome through coping, then how does it work? One finding that has relevance for the current study is that as compared with pessimists, optimists have fewer negative social interactions (Lepore & Ituarte, 1999). These authors believe that optimists’ more favourable outcomes are achieved through their self presentation to others. So, the caregivers who are more optimistic would have more positive interactions and less negative interactions because they present themselves in a positive way. These ensuing positive interactions are thought to help them adjust to the stressor. Clearly, this has implications for EE. It could be the case that caregivers who are high in DO present themselves more positively thereby encouraging more positive interactions and discouraging negative interactions, i.e. fewer interactions that comprise criticism or irritability. However, the relationship between DO and EE remains to be determined. A further possible explanation of how DO may work is given by Khoo and Bishop (1996) who suggest that it may be affecting the appraisal process “by looking on the bright side Optimists are likely to appraise the situation as being less threatening and hence less stressful” (p. 37). Chang (1998) suggests that optimism works more on secondary appraisal, i.e. in relation to the estimation of available resources. No studies relating optimism to the caregiving role in mental illness could be found, but in research on the caregiving role of adult daughters, Atienza, Stephens and Townsend (2002) found that dispositional optimism was beneficial in reducing the negative effects of stress. In light of the large amount of research supporting the role of coping in DO, the present study hypothesizes that DO is indeed associated with active coping, however, this relationship is thought to be mediated by the appraisal process as suggested by Khoo and Bishop (1996). 59 4.3 Cultural considerations 4.3.1 Chinese cultural implications for appraisal The cultural setting affects the appraisal process and Jenkins and Karno (1992) cite four features of EE that they believe are most affected by culture and which can be seen to impinge directly on the appraisal process: cultural beliefs about illness; cultural meanings of kin relations; cultural rule violations and vocabularies of emotion. Cultural beliefs about illness The question, “to what extent cultural conceptions of illness mediate EE in families?” (Jenkins & Karno, 1992, p.17) relates to the common held views about the nature and cause of mental illness in a particular culture. If the culture has a benign attitude to mental illness and views it as a legitimate illness then the individual will not have to contend with shame, stigma and so on. Also, people within a culture derive their own beliefs from those of the culture. Traditionally, the Chinese view of the etiology of mental illness is multifaceted, encompassing moral, religious, cosmological, physiological, psychological, social and genetic factors (Lin & Lin, 1980). However, a study by Lee & Bishop (2001) looking at Chinese beliefs (in Singapore) about the cause of psychological problems puts this in a more recent light. They found that the strongest endorsement was given to a psychological model with the indigenous model (Chinese medicine, Dang Ki, feng shui) receiving least support. So it may be the case in Singapore that the thinking is becoming more westernised than traditional Chinese, and that notions such as shame and face are less relevant. Vocabularies of emotion Relatives draw on important cultural knowledge of which emotions should or 60 should not be expressed in different situations; these are culturally prescribed ways of how one ‘should’ feel. Some cultures freely express anger and criticism whereas for others it is a taboo. As mentioned previously, shame is a common emotion in Chinese societies. Shame is one of a set of sub-conscious emotions that can be seen to be part of the appraisal process. If the situation is judged negatively, shame can occur and the individual reacts by either using avoidant coping or denial, or tries to blame others (Fischer & Tangney, 1985). If the patient is the recipient of this blame, it may lead to a hostile, critical atmosphere – high EE. It has also been shown to lead to anger, which again may contribute to high EE. It has been suggested that people in collectivist societies experience more shame than guilt compared to people in individualist societies and vice versa (Wallbott & Scherer, 1985). Therefore it is not surprising that the Chinese language has over 150 words for varieties of shame/guilt/embarrassment compared to the English language which has less than 30. Compared to the Americans and the Italians, the Chinese also see Shame as a basic emotion category, similar to love, anger, happiness, fear and sadness (Shaver et al.,1997, 1992 cited in Tangney & Fisher 1985). Cultural meanings of kin relations Cultures define what is the accepted way for families to interact and obviously the family is paramount in Chinese societies. As Lin & Lin (1980) state, the family controls the behavior of its members through a hierarchical structure and code of conduct. So for any Chinese family, faced with a stressor, preserving the family will be its priority. Examples of this have been seen in other cultures, for example, a study using Irish families found that patients with mental illness were ostracized as a way of keeping the moral standing of the family intact, an important factor for Irish families (Jenkins & Karno, 1992). So the importance of the family may also affect the way the relatives appraise and deal with the situation. 61 Violating cultural rules Different cultures have different rules regarding which behaviours are acceptable and which would be liable for criticism. For instance, in a Western country, lack of independence by a family member would be criticized whereas in Chinese societies, dependence is encouraged. In Singapore, Bentelspacher et al. (1994) noted that caregivers found the patient’s inability to work and bring money into the home a source of frustration. There may be other violations, perhaps in relation to filial piety, where children are not fulfilling their prescribed role in the family. 4.3.2 Chinese cultural Implications for coping In their study of people with Schizophrenia and their relatives in Singapore, Bentelspacher et al. (1994) found that the Singapore Chinese all used the coping strategy of Indifference-tolerance/resignation (unusual behaviour not perceived as a problem or accepted as unchangeable aspects of illness) most of the time, however these findings should be treated with caution, as the numbers were very small (n=11). These families were also found to lack informal supports and Bentelspacher et al. suggest that Chinese families emphasize self-directed coping strategies and rely on their immediate family when faced with a problem. Overall the Chinese families tended not to use helping networks or community mental health services. Bentelspacher et al., (1994) used Birchwood and Smith’s (1987) categories in their study, and although other coping measures (e.g. The COPE, Carver et al.,1989) have been used with non-Western samples, some Chinese psychologists (e.g. Cheng & Tang, 1993 cited in Cheng & Tang, 1995) argue that coping strategies used by Chinese people are not reflected in western scales, and highlight that whilst the coping styles are similar to those used by westerners, they include such resources as ‘appealing to supernatural power‘, acceptance, perseverance, praying to ancestors, etc. and not include emotional discharge (Taylor et al.,1992). 62 On the other hand, various researchers (e.g. Shek & Cheung, 1990 cited in Cheng & Tang, 1995) have identified coping styles that broadly fall into the ‘problem focused’ and ‘emotion focused’ coping styles of Folkman and Lazarus. In Taiwan, these were ‘relying on self’, ‘seeking help from social resources’, ‘appealing to supernatural power’ and ‘adopting a ‘do nothing’ philosophy’ (Hwang, 1977 cited in Cheng & Tang, 1995), similar categories were found in Hong Kong by Shek and Cheung, and by Cheng and Tang who interviewed 168 Chinese parents of children with Downs Syndrome. Marsella and Dash-Scheuer (1988), whilst accepting the possibility of universals in coping, stress that different cultures may have different conceptualizations of a problem. For example, in Chinese society, having a schizophrenic relative may not be as stressful as in Western societies. Cultural beliefs in the afterlife and concepts such as karma may affect perceptions of the illness. For example, the illness may have a positive aspect in that it may be regarded as a form of suffering which has come from a previous life (i.e. working off negative karma or debt) or which will impact positively on the next life (Lam & Palsane, 1996). A particular distinction between Western and Chinese cultures is that of Individualism and Collectivism (Hofstede, 1980). At the individual level, these distinctions are known as independent and interdependent self-construals (Markus & Kitayama (1991). The effects of interdependent self-construal have been found to influence cognitions and motivations (Markus & Kitayama, 1991). Of interest here is the effect that these self-construals may have on coping responses. It has been found that coping behaviours among Asians often involve family or community as sources of assistance (Kuo, Roysircar, & Newby-Clark, 2004). Kuo et al. (2004) using a new coping scale developed to measure collectivistic and individualistic dimensions of coping found three types of coping which were seen to correspond with Asian values. They were: Collective and avoidance coping approaches 63 (corresponding to family honour or filial piety); social harmony (corresponding to avoiding interpersonal conflict) and emotional control. 4.3.3. Chinese cultural implications for control attributions Morris and Peng (1994) hypothesize that attribution patterns reflect implicit theories acquired from socialization and that these will necessarily vary from culture to culture. The Chinese (a collectivist culture) tend to make situational or external attributions even in the case of negative situations whilst Individualistic cultures (e.g. US, UK, Europe) tend to make more dispositional or internal attributions. In this particular case, caregivers would make their attributions based on their knowledge or experience of mental illness and its causes. The question therefore is, to what extent belief models held in Singapore about the causes of mental illness differ from those held in the West. Whilst historically Singapore was considered a collectivist culture, it remains to be seen whether, given the large Western influences, it remains this way. As mentioned earlier, Lee and Bishop’s (2001) study on lay beliefs about the etiology and treatment of psychological problems among Singaporeans showed that the least endorsement was given to the indigenous beliefs model. The indigenous beliefs model contained Chinese medicine, dang-ki and feng-shui. The most endorsed model was the psychological model that contained psychodynamic, behavioural, humanistic and cognitive explanations. The author believed that this evolved from exposure to Western psychology through popular films, self-help books, pop psychology in popular magazine and seminars conducted by mental health professionals in additional to increased Westernisation in Singapore. It seems therefore, that belief models about causes of psychiatric illnesses held by Singaporeans would not differ greatly from those held by their Western counterparts. In addition, in a study conducted in China looking at attributions and expressed emotion, Yang (2003), found similar results to US/UK studies finding 64 support for the cross cultural validity of the proposed causal relation between attributions and EE. 4.3.4 Chinese cultural implications for Optimism Optimism is a cross-cultural construct and is one of the five basic emic dimensions of Chinese person perception (Cheng & Tang, 1995; Yang & Bond, 1990). According to Chinese folk wisdom, optimism is seen as being beneficial to health (Koo, 1987). Further, the relationship of stress-coping-well being has been demonstrated to have cross-cultural validity, Khoo and Bishop’s (1996) findings using a Singaporean sample were congruent with studies using North American samples. In Hong Kong, Lai and colleagues (Lai, Cheung, Lee, & Yu, 1998) have found results that are consistent with previous findings among western samples. In addition, scores on the LOT were found to moderate the relation between hassles and somatic complains (Lai, 1995) showing validity for the scale in this population. Optimistic students were found to use more adaptive strategies than their less optimistic peers (Lai & Wan, 1996). Chang (2001) found no differences between White and Asian American students on levels of optimism. Similarly, Ji, Zhang, Usborne and Guan (2004) found no significant cultural differences on DO as measured by the LOT-R. Taken together, these findings support the applicability of the concept of optimism and the LOT in particular, to research on optimism in Chinese populations. 4.4 Summary To sum up, there are inconsistencies in the predictive power of EE for first- episode and Asian populations. The usefulness of self-report measures is highlighted and the respective or combined roles of DUI and EE in relation to predicting outcome needs to be clarified. 65 With regards to the environmental protectors, there is evidence that caregivers higher in EE also appraise the situation more negatively, perceive the patient’s symptoms more negatively, use more avoidant coping and hold more internal attributions of illness control. Dispositional optimism is hypothesized to have an influence on Active Coping through appraisal. In addition there are many cultural factors in relation to these factors that must be considered. 66 Chapter 5: Potential predictors of outcome in psychosis – Duration of untreated illness 5.1 Duration of Untreated Psychosis and Duration of Untreated Illness Duration of untreated illness is the time from when an individual starts to have symptoms of the illness until they see a psychiatrist. Researchers (e.g. Lincoln & McGorry, 2000) look at this delay in seeking treatment from two slightly different perspectives: the duration of untreated illness (DUI), which is the time from the start of the prodromal period to contact with a psychiatrist and the duration of untreated psychosis (DUP) which is the length of time from when the patient first experiences psychotic symptoms (e.g. hallucinations, delusions etc) and first contact with a psychiatrist (see figure 5.1). Prodromal Phase of Psychosis DUI Acute Phase Of Psychosis DUP Meeting with Psychiatrist Figure 5.1: Duration of Untreated Illness (DUI) and Duration of Untreated Psychosis (DUP) in relation to Prodromal and Acute phases of illness 67 This period of untreated illness, the critical period, is not only a time when neurological damage may occur; as Birchwood (2000, p. 246) states, “biological psychosocial and cognitive changes which are influential in the course of psychosis are not ‘given’ but actively develop during this period”. Birchwood suggests that the patient’s appraisal of the situation (which seems to be firmly established in this early period) is one of lack of power over his or her illness, and that this feeling of entrapment can often stop the patient from moving on and establishing an identity. Early intervention therefore also aims to help the patient at this time to preserve a sense of self, empower themselves and increase self-efficacy. There can be many reasons for this delay in seeking treatment. With regard to DUI, the prodromal signs can often go unnoticed because many of the behaviours can be attributed to ‘normal’ adolescence such as withdrawal, neglect of personal hygiene, etc (Hulbert, Jackson & McGorry, 1996). With regard to DUP, delays may be due to other causes. For instance, Singapore has a much longer average DUP than other Western countries (Chong et al., 2005) and this may be because families may seek help initially from traditional healers. Evidence for this comes from Bentelspacher et al. (1994) who found that many Singaporean families consulted traditional healers before approaching a psychiatrist, either bomohs (Malay), temple mediums (Chinese or Indian) or herbalists (Chinese). A more recent study confirms these findings, in a recent survey on a patient population in Singapore it was found that 28% went to traditional healers first whilst only 21% consulted a psychiatrist first (Choo et al., 2003). Another reason for delay in seeking treatment is due to the stigma attached to having a relative with mental illness (Chan, 1995). Also, depending on the diagnostic criteria used (e.g. ICD-10 World Health Organisation; DSM IV American Psychiatric Association 1994), patients would have to display psychotic symptoms from one to six months before being formally diagnosed as having schizophrenia and treatment initiated. 68 Numerous centres around the world are now dedicated to the study of early psychosis, focusing on individuals in the initial stages of schizophrenia or a related disorder, and their caregivers (e.g. European First Episode Schizophrenia network (EFESN) 1995; Early Psychosis prevention and intervention centre (EPPIC), Melbourne 1992; Early Psychosis Intervention Programme (EPIP), Singapore). 5.2 The Association of DUP and DUI with Outcome Unlike the research investigating the relationship of EE and Outcome, researchers in the area of DUP and DUI have looked at a variety of outcome measures. Macmillan et al. (1986) were one of the first to notice the relationship between longer DUI and poorer outcome, and to raise the point that DUI may be predicting outcome, not EE. This controversy over which of EE and DUI is the better predictor is one that needs to be addressed. However as mentioned previously, much of the research conducted in this regard has been the subject of methodological or statistical criticism and studies have often used very small sample sizes which reduces confidence in the findings (see Stirling et al., 1991). 5.2.1 DUP There is enormous support for DUP as a predictor of outcome. Studies have found DUP to be a predictor of remission (e.g. Loebel et al., 1992), positive symptoms (e.g. Larsen, Moe, Vibe-Hansesn & Johannessen, 2000) and general symptoms (Bottlender et al., 2002; Larsen et al., 2000) In summarizing the literature, Norman, Lewis and Marshall (2005) report that of ten studies that looked for a relationship between DUP and either remission or positive symptoms, nine found statistically significant relationships between longer DUP and lower remission rates or higher positive symptom ratings. There are also a few studies which have failed to find a relationship between DUP and outcome (e.g. Ho, Andreasen, Flaum, 69 Nopoulos & Miller, 2000; Barnes, Hutton, Chapman, Mutsatsa & Joyce, 2000; Craig et al., 2000). However, Harrigan, McGorry and Krstev (2003) suggest that in Ho et al. (2000) and Barnes et al. (2000) studies the sample sizes (N = 74 and N = 53 respectively) may have been too small to detect the small to moderate effect sizes which are typical of the DUP and outcome relationship. The situation with regards to negative symptoms is less clear. Although some studies have found DUP to be associated with negative symptoms (e.g. Oosthuizen, Emsley, Keyter, Niehaus & Koen, 2004). Norman et al. (2005) report that of ten studies, only four found a statistically significant relationship between longer DUP and more negative symptoms. Looking now at non-symptomatic outcome measures, longer DUP has also been associated with poorer social functioning at one year (Larsen et al., 2000; Addington, van Mastrigt & Addington, 2004) and with lower measures of QOL at one year (Addington, Young & Addington, 2003; Addington et al.,2004; Browne et al., 2000; Harrigan et al.,2003; Malla et al., 2004). There have been suggestions (e.g. Verdoux et al., 2001) that the effect of DUP on outcome is spurious and that the relationship could be explained by other factors such as age at onset, type of onset, gender and particularly, premorbid adjustment. To clarify this situation, a number of studies have examined the relationship between DUP and outcome while controlling for these factors and have concluded that DUP is an independent predictor of outcome (Addington et al., 2004; Bottlender et al., 2002; Harrigan et al., 2003; Larsen et al., 2000; Loebel et al., 1992; Malla et al., 2002). So clearly, DUP has been shown to be an independent and robust predictor of symptomatic and functional outcome. 70 5.2.2 DUI DUP has been the focus of research more than DUI. This is primarily because DUP involves easily recognisable symptoms such as hallucinations or delusions which make it easier to identify and target. DUI on the other hand involves prodromal symptoms which are less easily identifiable and are often mistaken for normal adolescence. In this way, the prodromal period is usually over by the time the patient sees the psychiatrist. However, whilst DUI has received less attention, it is still nonetheless an important factor. DUI has been found to predict symptomatological outcome (Loebel et al., 1992; Rabiner Wegner & Kane ,1986). In addition, whilst Hafner et al. (1998) found that DUP was a significant predictor only of psychotic and non-specific symptoms at five year follow up, DUI predicted negative and non specific symptoms and social outcome. These authors point out that as the period of DUP is characterized by positive symptoms and that of DUI characterized by negative symptoms, it makes sense that they should predict those symptoms. Keshavan et al., (2003) found DUI to be a highly significant predictor of social and occupational functioning, even after controlling for pre-morbid adjustment and Browne (1996) found DUI to be related to QOL. So clearly, DUI is also an important factor regarding prediction of outcome in different domains. 5.3 Summary DUP and DUI are both important in terms of their relationship with outcome. It seems that these factors are especially important in this population which has longer periods of untreated illness than Western populations. 71 Chapter 6: 6.1 Outcome Definitions of outcome used in previous studies As mentioned in the introduction, the focus now in terms of outcome is on a multi-dimensional approach incorporating symptomatic and functional measures. Previously however, outcome was not a clearly defined measure and relapse definitions themselves varied greatly over the years since the inception of EE. In the early British studies (Brown et al., 1972) relapse was defined as a change from normal or non-schizophrenic state to a state of schizophrenia identified through the Present State Examination (PSE; Wing, Cooper & Sartorius, 1975) showing a marked exacerbation and persistent schizophrenic symptoms. ‘No relapse’ was defined as no schizophrenic symptoms at follow-up or symptoms remaining steady. In the Californian replication of the early British studies (Vaughn et al., 1984), relapse was defined as an increase of three points on one or more of the scales of the Premorbid Adjustment Scale (PAS; Cannon-Spoor, Potking & Wyatt, 1982): hallucinations, delusions and incoherent speech. Not all studies have used such precise outcome measures, several studies used readmission, for any reason (Macmillan et al., 1986), following exacerbation of clinical features (Parker et al., 1988; Stirling et al., 1991) or for showing typical positive symptoms (Bertrando et al., 1992). Others use marked increases in the Brief Psychiatric Rating Scales (BPRS; Overall & Gorham, 1962): (Ng et al., 2001, Nuechterlein, Snyder & Mintz, 1992) or Schedule for Affective Disorders and Schizophrenia scales (SADS-L; Endicott & Spitzer, 1978)(Moline et al., 1985). Some based relapse on examination of clinical records (Linszen et al., 1996) or included suicide attempts (Phillips & Xiong, 1995). This variation in relapse measures suggests that the ability of EE to predict relapse may be over, or under-inflated. For instance, King and Dixon (1999) studied 69 patients with chronic schizophrenia. Using conservative criteria for relapse 72 (readmission, increase in antipsychotic medication of 100% or more), EE failed to predict and 12 month relapse. Using standard criteria (readmission, increase in antipsychotic medication of 50% or more) they found that month relapse rates were significantly greater among patients in high EE homes. 6.2 Outcome from a broader perspective As mentioned, there is a general move away from purely symptomatological outcome measures. For instance, Drake, Haddock, Hopkins and Lewis (1998) believe that other measures, such as social functioning or quality of life (QOL) should be included. With regards to the former, it may be that better social functioning in a patient may indicate better social skills which may in turn lead to more social interaction and less likelihood of criticism and hostility from the family. With regard to QOL, patients from low EE homes, although they not suffer relapse, may have reduced quality of life. They may have become so withdrawn that their behaviours not provoke criticism, nor they lead to relapse, so in this case a low EE family environment may not be as benign as first thought. Alternatively, patients from a high EE family may have good quality of life and relapse may be due to some other factor. Liberman et al. (2002) also argue that by using only the presence or absence of psychiatric symptoms, a complete representation of the patient’s outcome is missed, for example, positive symptoms which occur during the patient’s follow-up appointment may be brief and have little impact on their social or occupational functioning. Oliver (1999) sees this move towards using QOL as being an extremely important advance in evaluating outcome in mental health, moving away from a narrow symptom-centred view to one that is more holistic. Some concern has been raised over the ability of patients with psychosis to accurately rate their own quality of life given that they may be suffering from disturbed thinking. However, the consistency and reliability of patient reports have 73 been reported in several recent studies (Awad, Voruganti, & Heselgrave, 1995; Voruganti, Heselgrave, Awad, & Seeman, 1998). Further, Herrman, Hawthorne and Thomas (2002) suggest that patient self-reports should be the preferred option, given the systematic differences in patient and objective reports. The outcome measures will be defined and described fully in Part 2. 6.3 Summary A broader multi-dimensional view of outcome incorporating functional as well as symptomatological measures and including QOL is now recommended and this is adopted in the present study. Continuous measures rather than a dichotomous relapse/non-relapse distinction also provide a more detailed picture of the patient’s outcome. 74 [...]... unusual to find this combination in a Western population Ran et al (2003, p 10 3) quote a Chinese saying, ‘the deeper the love, the greater the criticism’, which sums up the meaning of criticism for the Chinese 16 However, to what extent would these traditional Chinese cultural ideas be found in Singapore, given its Western influences? A recent study on the Singapore family conducted using Singaporean... harshness with inhibition of the expression of opinions (Ho, 19 96; Minuchin, Rosman & Baker 19 78, cited in Hsu, 19 95) These somewhat strict parenting behaviours, together with the acceptance of criticism, may render some of the factors within the EE concept as normal behaviour in this population 1. 6.3 Extended Family The lack of EOI in the Indian sample mentioned previously led Leff et al (19 87) to suggest... since the end of WWII traditional Chinese practices, customs and values have been on the decline in Singapore There are several factors cited for this decline, education being one English education was not adopted by many Chinese speakers in the early years, but after WWII, it began to increase and overtook enrolment in Chinese schools by 19 54 (Tsoi, 19 85), with a further shift after Singapore gained... remains to be seen how these factors will affect the present study 1. 6 .1 Emotional over-involvement (EOI) Different rates of EOI found cross-culturally suggest that this concept may have different connotations in non-Western society For instance, In Leff et al’s (19 87) study in India, no family was found to score high on EOI By comparison, 11 % of Mexican descent, 15 % of Anglo American and 21% of the... EOI rates in Bali where many family members live in one household But, many relatives also live in the same compound, suggesting that family support is protective In Singapore, the extended family is on the decline as the proportion of single family households has risen steadily from 63.5% in 19 57 to 17 85% in 19 90 The government however, recognizing the importance of the extended family, has introduced... (19 91) in Finland Tienari and colleagues compared the psychiatric status of babies of schizophrenic mothers who were adopted, with babies of non-psychiatric mothers who were also adopted Of the adopted children, 15 turned out to be psychotic, 13 of these were from the schizophrenic mothers clearly showing the genetic link In addition, they found that the adopted families of the 15 psychotic offspring... Singapore gained independence in 19 65 At the same time as the Chinese were moving away from Chinese culture and towards the English education system, many also converted to Christianity, causing further decline away from traditional practices, customs and values (Kuo, 19 85) 14 Tsoi (19 85) sees Singaporeans as constituting two main groups: An older generation, born mostly before WWII, very Chinese in nature... of the EE index: EOI and Criticism A further two are related to cultural factors which may affect levels of EE: the extended family, and the importance of the family However, before looking in more detail at these factors, another issue which must be borne in mind is the extent to which Singaporean Chinese still adhere to traditional Chinese customs and ways of thinking and acting Researchers are in. .. disturbances in the mind and/or brain The person with psychosis will suffer a range of disturbed cognitions, the main ones are now discussed (from Choo, Verma & Chong, 2003): Disturbed perceptions: Disturbed perceptions, or hallucinations, can affect any of the senses Hearing voices, seeing images, feeling things touching the body, having strange tastes in the mouth or smelling things that aren’t there in reality... experienced an episode of psychosis Similarly, Northern Indian relatives are much 13 more accepting of schizophrenia developing in a family member than relatives in England or the USA (Lefley, 19 89) Actual EE ratings reflect these different attitudes; with rates of EE varying considerably cross-culturally For instance, the following are rates of high EE found in four different cultural groups: Indians (23%), . multi-cultural society comprising Chinese (77%), Malay (14 %), Indian (7%) and other (1. 7%) (Statistics Singapore, 2003). As the majority of the population in Singapore is of Chinese background, this. in non-Western society. For instance, In Leff et al’s (19 87) study in India, no family was found to score high on EOI. By comparison, 11 % of Mexican descent, 15 % of Anglo American and 21% of. with inhibition of the expression of opinions (Ho, 19 96; Minuchin, Rosman & Baker 19 78, cited in Hsu, 19 95). These somewhat strict parenting behaviours, together with the acceptance of criticism,