- Email: [email protected]
Differentiating schizoaffective and bipolar I disorder in first-episode psychotic mania
Schizophrenia Research 140 (2012) 31–36
Contents lists available at SciVerse ScienceDirect
Schizophrenia Research journal homepage: www.elsevier.com/locate/schres
Differentiating schizoaffective and bipolar I disorder in first-episode psychotic mania Daniel Schöttle a,⁎, 1, Benno G. Schimmelmann b, 1, Philippe Conus c, Sue M. Cotton d, Chantal Michel b, Patrick D. McGorry d, Anne Karow a, Dieter Naber a, Martin Lambert a a Psychosis Early Detection and Intervention Centre (PEDIC), Centre for Psychosocial Medicine, Department for Psychiatry and Psychotherapy, University Medical Centre Hamburg-Eppendorf, Germany b University Hospital of Child- and Adolescent Psychiatry, University of Bern, Switzerland c Treatment and early Intervention in Psychosis Program (TIPP), Department of Psychiatry CHUV, Lausanne University, Switzerland d Orygen Youth Health Research Centre, Centre for Youth Mental Health, University of Melbourne, Parkville, Australia
a r t i c l e
i n f o
Article history: Received 27 February 2012 Received in revised form 25 June 2012 Accepted 9 July 2012 Available online 28 July 2012 Keywords: Schizoaffective disorder Bipolar I disorder First-episode
a b s t r a c t Objective: This study aims to differentiate schizoaffective disorder (SAD) and bipolar-I-disorder (BD) in first-episode psychotic mania (FEPM). Methods: All 134 patients from an epidemiological first-episode psychosis cohort (N=786) with FEPM and an 18-month follow-up final diagnosis of SAD (n=36) or BD (n=98) were assessed with respect to pre-treatment, baseline and outcome differences. Second, patients with baseline BD who shifted (shifted BD) or did not shift to SAD (stable BD) over the follow-up period were compared regarding pre-treatment and baseline differences. Results: SAD patients displayed a significantly longer duration of untreated psychosis (DUP; effect size r=0.35), a higher illness-severity at baseline (r=0.20) and more traumatic events (Cramer-V=0.19). SAD patients displayed a significantly higher non-adherence rate (Cramer-V=0.19); controlling for time in treatment and respective baseline scores, SAD patients had significantly worse illness severity (CGI-S; partial η2 =0.12) and psychosocial functioning (GAF; partial η2 =0.07) at 18-months, while BD patients were more likely to achieve remission of positive symptoms (OR=4.9, 95% CI=1.8–13.3; p=0.002) and to be employed/occupied (OR=7.7, 95% CI= 2.4–24.4, p=0.001). The main discriminator of stable and shifted BD was a longer DUP in patients shifting from BD to SAD. Conclusions: It is difficult to distinguish BD with psychotic symptoms and SAD in patients presenting with FEPM. Longer DUP is related to SAD and to a shift from BD to SAD. Compared to BD, SAD had worse outcomes and higher rates of non-adherence with medication. Despite these differences, both diagnostic groups need careful dimensional assessment and monitoring of symptoms and functioning in order to choose the right treatment. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Despite the ongoing controversy on the diagnostic entity of schizoaffective disorder (SAD) and its distinction from bipolar I disorder (BD) or schizophrenia (Cheniaux et al., 2008, 2009), it is a clinical reality that a subgroup of patients with psychosis suffers from a combination of mood unrelated psychotic episodes and alternating depressive and manic syndromes with psychotic symptoms. Compared to patients with schizophrenia or bipolar I disorder, such a combination may need specific treatment including pharmacotherapy and psychosocial interventions (Murru et al., 2011).
⁎ Corresponding author at: Psychosis Early Detection and Intervention Centre (PEDIC), Centre for Psychosocial Medicine, Department for Psychiatry and Psychotherapy, University Medical Centre Hamburg-Eppendorf, Martinistrasse 52, 20246 Hamburg, Germany. Tel.: +49 152 22816853; fax: +49 40 741055455. E-mail address: [email protected] (D. Schöttle). 1 Denotes equal contribution to the study. 0920-9964/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.schres.2012.07.010
However, in first-episode psychotic mania it is difficult to foresee whether a patient will eventually have SAD or BD or, specifically, if initially diagnosed with BD, whether the patient will shift to SAD during the course of treatment. In the first and only study on this topic so far, Conus et al. (2010a) reported the following differences between SAD and BD: lower premorbid functioning level, longer duration of prodrome and untreated psychosis, higher severity of positive symptoms at first presentation and, at 12 months follow-up, more severe negative symptoms and a poorer functional level. This study, however, has one important limitation: data were not derived from a representative cohort; this suggests that the inclusion of patients presenting with a high level of severity of both psychotic and manic syndromes and high rates of comorbidities (e.g. substance use disorders) may have been limited. 1.1. Aims of the study The aims of this study were to (1) differentiate patients presenting with first-episode psychotic mania (FEPM) with a discharge diagnosis
32
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
of schizoaffective disorder (SAD) and those with a discharge diagnosis of bipolar I disorder (BD) in terms of pre-treatment, baseline and outcome characteristics and (2) assess pre-treatment and baseline differences between FEPM patients with baseline BD who shifted (=shifted BD) or did not shift to SAD (=stable BD) within an 18-month follow-up period. 2. Materials and methods 2.1. Context and sample The Early Psychosis Prevention and Intervention Centre (EPPIC) is a comprehensive program for young people aged 15–29 years experiencing their first treated episode of psychosis. Treatment spans an average of 18 months. The EPPIC catchment area covers the northwestern regions of Melbourne, Australia. There is a lack of other facilities for the target population in these regions and a scarceness of private psychiatrists. Leakage to private facilities outside the catchment is rare. EPPIC therefore ascertains a treated epidemiological sample of FEP patients (McGorry et al., 1996). The First Episode Psychosis Outcome Study (FEPOS) is a file audit study of all 786 first-episode patients who were admitted to EPPIC between 1998 and 2000 (Lambert et al., 2005b; Conus et al., 2007; Schimmelmann et al., 2007, 2008; Cotton et al., 2009; Robinson et al., 2009; Conus et al., 2010a, 2010b; Robinson et al., 2010; Schimmelmann et al., 2011a, 2011b; Cotton et al., 2012). Of the 786 patients admitted, eighty-two (10%) of the patients' files had been transferred to other services and 43 (5%) patients were excluded because they had a non-psychotic diagnosis at discharge. Of the remaining 661 patients, 182 (27.5%) had a final 18-month diagnosis of SAD or BD. Patients initially presenting with a first-episode manic– psychotic syndrome (regardless of initial psychosis diagnosis) were selected. To ensure the inclusion of clearly manic patients, only those who fulfilled the criterion of at least ≥4 points (moderately to very severely manic) in the mania subscore of the Clinical Global Impression – Scale for use in BD (CGI-BP; Spearing et al., 1997) were included. Data of 134 patients (73.6%) were analyzed, 98 (73%) of those received a final diagnosis of BD and 36 (27%) with a final diagnosis of SAD. 2.2. Assessments and measures For each patient treated at EPPIC, information on pre-treatment, baseline (admission to EPPIC), treatment and outcome characteristics are systematically documented in a structured file. Assessments are based on the Royal Park Multi-diagnostic Instrument for Psychosis (RP-MIP; McGorry et al., 1990a, 1990b). Each file contains information compiled during the 18 month treatment period from various sources using high quality assessments carried out by trained clinicians. Two experienced psychiatrists assessed all files using a standardized questionnaire (Early Psychosis File Questionnaire, EPFQ; Conus et al., 2007). Pre-treatment-, baseline-, treatment and outcome-variable: Premorbid functioning, was assessed with the GAF (American Psychiatric Association, 1994). Baseline-, treatment and outcome-variables: Severity of illness with the CGI-S (Guy, 1976); employment/occupation (yes/no) with the MVCI (Tohen et al., 2000a, 2000b); lack of insight into illness on the basis of one item with anchors ranging from lack of insight to partial and full insight (Conus et al., 2007). Pre-treatment variables: gender; family history of psychosis (yes/no) rated as present in a first or second degree relative (Morley et al., 2008); traumatic history (yes/no) refers to sexual abuse (sexual molestation and/or rape) and/or physical abuse (physical attack or assault or being repeatedly beaten by parents, relatives, or caregivers during childhood (Conus et al., 2009, 2010b)); criminal justice history (yes/no) was considered if there was mention in the file of previous contact with the legal system for any form of conviction; age at onset of psychosis was defined as the age when one first sustained positive psychotic
symptoms occurred according to the DUP scale (McGorry et al., 1990a, 1990b); duration of untreated psychosis with the DUP scale, for detailed procedure see Schimmelmann et al. (2008); lifetime SUD with the DAAS (Lambert et al., 2005a); for detailed procedure see Lambert et al. (2005a); past suicide attempts according to ICD-10 classification (Dilling and Dittmann, 1990). Baseline variables: severity of mania with the CGI-BP (Spearing et al., 1997); age; living with family (yes/no) with the MVCI (Tohen et al., 2000a, 2000b); and co-morbid SUD with the DAAS (Lambert et al., 2005a). Treatment and outcome variables: Time in treatment was calculated on the basis of the regular 18-month treatment in EPPIC and, in the case of service disengagement, the time point when a patient dropped out of treatment; service disengagement was defined as present if patients actively refused any contact with the treatment facility or were not traceable (for detailed procedure see Conus et al., 2010c and Schimmelmann et al., 2006); persistent SUD with the DAAS (for detailed procedure see Lambert et al., 2005a); medication non-adherence was defined according to Robinson et al. (2002) as failure to take medication for 1 week or longer (for detailed procedure see Lambert et al., 2010); suicide attempts in treatment according to ICD-10 classification (Dilling and Dittmann, 1990); hospital admission (yes/no); remission of positive symptoms (yes/no) was defined as receiving a score of no worse than “mild” (score≤ 3) in the CGI-S on discharge from the service or time of service disengagement (according to Lambert et al., 2008). 2.3. Diagnostic assessment, validity, and inter-rater reliability Clinical diagnoses according to DSM-IV criteria (American Psychiatric Association, 1994) at EPPIC are the consensus result of an intensive diagnostic and treatment process, first within the initial 6 weeks of admission by well-trained clinicians working in a specialized assessment and crisis assertive community treatment team, and then at discharge based on all available information. The discharge diagnoses were used to differentiate SAD and BD in this study. Diagnoses were extracted from the files by the principal investigators (ML and PC). In case of disagreement with clinical diagnoses reported in the file, a consensus rating between both research psychiatrists and the patient's case manager was performed (Schimmelmann et al., 2005). Validity of the FEPOS diagnoses was established by the following procedure: between 1998 and 2000, 230 of the 786 patients treated at EPPIC have been included in prospective trials. Their main and co-morbid diagnoses were assessed within 6 weeks of admission using the Structured Clinical Interview for DSM-IV (SCID-I/P; Ventura et al., 1998). The randomly selected SCID and FEPOS diagnoses of 115 patients were compared. The calculated kappa values revealed a very good concordance for both psychosis diagnoses (kappa=0.80) and co-morbid substance use disorder (SUD) diagnoses (kappa=0.74). Inter-rater reliability has been established by comparing baseline ratings given independently by both main investigators in a randomly selected sample of 40 files stratified by time, on the following scales: CGI, GAF, and insight. Analysis revealed a good to very good inter-rater reliability with kappa values ranging from 0.80 to 0.90 (CGI-S: 0.87, GAF score: 0.88, insight score: 0.89). 2.4. Data analysis Patients with discharge SAD and BD were compared with respect to pre-treatment, baseline, and treatment characteristics using Mann– Whitney U-tests when the dependent variable was continuous and chi square analysis (χ2), when the dependent variable was categorical. With respect to outcome differences in terms of CGI-S and GAF, two one-way analyses of covariance (ANCOVAs) were specified controlling for time and treatment and the respective baseline scores. For differences in remission of positive symptoms and employment/occupation rates, two logistic regression analyses were specified with time in service and the respective baseline values (CGI-S and employment/occupation)
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
were entered into the first step, and the final diagnosis of SAD or BD were entered into the second step. For these analyses, the odds ratios (OR) and the 95% confidence interval (CI) of the ORs were calculated and the Wald statistic (z) was used to determine the significance of the associations (only p-values are reported for these analyses). To estimate effect sizes, Cramer-V was used for χ2 tests, the formula r= Z/√(N) for Mann–Whitney U-tests, and partial η2 for ANCOVAs. For χ2 statistics post-hoc examination of standardized residuals (z) was applied to determine which cells contributed to the significant χ2 if appropriate (only p-values will be reported). Pre-treatment and baseline variables of shifted and stable BD were compared applying the above mentioned statistical techniques. The retrospective consistency of the two diagnostic categories was calculated, (i.e., the proportion of patients in a discharge diagnostic category who received the same diagnosis at baseline). All analyses were performed using the IBM® Statistical Package for Social Sciences version 17 (IBM® SPSS® 17.0; Chicago, IL). 3. Results The sample characteristics and differences between patients with a final diagnosis of schizoaffective (SAD) and bipolar I disorder (BD) are displayed in Tables 1 and 2. 3.1. Pre-treatment and baseline differences Patients with SAD had a higher overall rate of traumatic events (44% vs. 25%; p = 0.025), which was due to a higher rate of physical abuse (41% vs. 20%; p = 0.022). Compared to BD, SAD patients were found to have a longer overall DUP (mdn. 31 days vs. 0 days, p b 0.001); specifically the rate of a very short DUP (1–30 days; 42% vs. 69%) was lower in SAD and a long DUP (91–365 days; 31% vs. 4%) was higher in SAD. At baseline, SAD patients had a higher illness severity (CGI-S 6.1 vs. 5.8, p = 0.022). However, the effect sizes for pre-treatment and baseline differences were low except regarding DUP (effect size= 0.35). 3.2. Treatment and outcome differences Patients with SAD stayed in treatment for a significantly longer time (73.0 vs. 59.1 weeks; p =0.041) and were more likely to be nonadherent with medication (72% vs. 51%; p =0.030; see Table 2). No differences were detected regarding service disengagement, persistent substance use, suicide attempts in treatment and rate of patients being hospitalized. Controlling for time in treatment and the respective baseline scores, SAD patients had higher illness severity and worse psychosocial functioning at discharge (adjusted means= 3.2 vs. 2.2; pb 0.001 and 61.5 vs. 69.8; p= 0.002, respectively). The effect sizes for these outcome differences were small. However, controlling for time in treatment and baseline CGI-S, BD patients were clearly more likely to achieve remission of positive symptoms at discharge (OR= 4.9, 95% CI= 1.8–13.3; p = 0.020) and to be employed/occupied (OR= 7.7, 95% CI =2.4–24.4, p= 0.001) compared to SAD patients. 3.3. Pre-treatment and baseline differences between stable and shifting BD In this FEPM sample, the retrospective diagnostic stability of SAD was lower than that of BD (41% vs. 92%; p b 0.001; effect size= 0.68; see Table 3). 33% of patients with a final diagnosis of SAD were initially diagnosed with BD. We further explored whether patients initially diagnosed with BD, who subsequently shifted to SAD (n= 12; shifting BD), differed with regard to pre-treatment and baseline variables from those, who consistently received a BD diagnosis at baseline and discharge (n= 90, stable BD). In other words, it is possible to predict a diagnostic shift to SAD in those, who initially present with BD with psychotic symptoms. As the sample size is low, effect size was chosen as the main statistical
33
Table 1 Pre-treatment and baseline differences between patients with schizoaffective and bipolar disorder. Diagnosis Schizoaffective (n = 36; 26.9% ) Pre-treatment characteristics Male, n (%) 23 (63.9%) Premorbid GAF 71.4 (9.6) score, M (SD) 9 (25.0%) Family history of psychosis, n (%) Traumatic 16 (44.4%) history, n (%) Sexual abuse 8 (22.2%) Physical 13 (40.6%) abuse Criminal justice 9 (26.5%) history, n (%) Age at onset of 22.6 (3.7) psychosis, M (SD) 31 (28–123) DUP, Mdn (quartiles), days DUP-categories 1–30 days 15 (41.7%) 31–90 days 8 (22.2%) 91–365 days 11 (30.6%) >365 days 2 (5.6%) Lifetime SUD, 31 (86.1%) n (%) 3 (8.3%) Past suicide attempts, n (%) Baseline characteristics CGI-S score, 6.1 (0.6) M (SD) CGI-BP mania, 5.1 (0.9) M (SD) GAF score, 26.3 (7.7) M (SD) Age at baseline, 23.0 (3.6) M (SD) in years 19 (52.8%) Employment/ occupation, n (%) 23 (63.9%) Living with family, n (%) Lack of insight, 28 (77.8%) n (%) Co-morbid 27 (75.0%) SUD, n (%)
Total
Statistics
Bipolar I (n = 98; 73.1%)
(N = 134)a
p-value Effect b sizec
58 (59.2%) 74.4 (10.8)
81 (60.4%) 73.6 (10.5)
0.621 0.094
0.04 0.14
26 (28.3%)
35 (27.3%)
0.710
0.03
24 (24.5%)
40 (29.9%)
0.025
0.19
16 (16.3%) 16 (19.8%)
24 (17.9%) 29 (25.7%)
0.430 0.022
0.07 0.22
19 (19.6%)
28 (21.4%)
0.400
0.07
0.462
0.06
b0.001
0.35
22.1 (3.2)
0 (0–31)
68 23 4 3 73
(69.4%) (23.5%) (4.1%) (3.1%) (74.5%)
4 (4.1%)
22.3 (3.3)
30 (0–31)
b0.001
0.39
83 (61.9%) 31 (23.1%) 15 (11.2%) 5 (3.7%) 104 (77.6%)
0.153
0.12
7 (5.3%)
0.334
0.08
5.8 (0.7)
5.9 (0.7)
0.022
0.20
5.3 (0.9)
5.2 (0.9)
0.226
0.10
30.0 (9.2)
29.0 (8.9)
0.094
0.18
22.4 (3.1)
22.5 (3.2)
0.360
0.08
67 (68.4%)
86 (64.2%)
0.095
0.14
48 (49.5%)
71 (53.4%)
0.139
0.13
76 (78.4%)
104 (78.2%)
0.943
0.01
61 (62.2%)
88 (65.7%)
0.168
0.12
Bold letters indicate statistically significant differences between SAD and BD (p b 0.05). Abbreviations: GAF = Global Assessment of Functioning Scale, DUP = duration of untreated psychosis, SUD = substance use disorder, CGI = Clinical Global Impression — Severity of Illness Scale, M = mean, Mdn= median, SD = standard deviation. a Ns vary due to missing data. b Descriptive statistics: p-values χ2 for categorical variables, Mann–Whitney U-tests for continuous pre-treatment and baseline characteristics. c Effect size = Cramer-V for χ2 tests and Z/√(N) for Mann–Whitney U-tests.
descriptor. Only differences with an effect size at least approaching moderate level, i.e. of at least 0.25, were considered indicative of a noteworthy group effect. Only DUP-categories were different between stable and shifting BD: the rate of a very short DUP (1–30 days; 50% vs. 69%) was lower in stable BD and a long DUP (91–365 days; 33% vs. 4%) was higher in shifting BD (Fisher-test; χ2 =12.4; p =0.022; effect size= 0.35).
34
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
Table 2 Treatment and outcome differences between patients with schizoaffective and bipolar I disorder. Diagnosis
Time in treatment, M (SD) weeks Service disengagement, n (%) Persistent SUD, n (%) Medication non-adherence, n (%) Suicide attempts in treatment, n (%) Hospital admission, n (%) Remission of positive symptoms, n (%) Employment/occupation d, n (%) CGI-S score e, M (SD) GAF score e, M (SD) Lack of insight, n (%)
Total
Statistics
Schizoaffective (n = 36; 26.9% )
Bipolar I (n = 98; 73.1%)
(N = 134)a
p-valueb
Effect sizec
73.0 (27.6) 4 (11.1%) 8 (22.2%) 26 (72.2%) 4 (11.1%) 35 (97.2%) 22 (61.1%) 6 (22.2%) 3.2 (1.3) 61.5 (13.6) 5 (13.9%)
59.1 (33.7) 20 (20.4%) 19 (19.4%) 47 (51.1%) 6 (6.1%) 88 (89.8%) 79 (80.6%) 50 (60.2%) 2.2 (1.2) 69.8 (14.2) 6 (6.1%)
62.9 24 27 73 10 123 101 56 2.5 67.4 11
0.041 0.213 0.717 0.030 0.330 0.165 0.020 0.001 b0.001 0.002 0.147
0.18 0.11 0.03 0.19 0.08 0.12 0.20 0.33 0.12 0.07 0.13
(32.7) (17.9%) (20.1%) (57.0%) (7.5%) (91.8%) (75.4%) (50.9%) (1.3) (14.4) (8.2%)
Bold letters indicate statistically significant differences between SAD and BD (p b 0.05). Abbreviations: GAF = Global Assessment of Functioning Scale, DUP = duration of untreated psychosis, SUD = substance use disorder, CGI = Clinical Global Impression — Severity of Illness Scale, M = mean, Mdn = median, SD = standard deviation. a Ns vary due to missing data. b Statistics: p-values χ2 for categorical variables, Mann–Whitney U-test for time in treatment, ANCOVA for CGI-S and GAF controlling for time in treatment and the respective baseline values. c Effect size = Cramer-V for χ2 tests, Z/√(N) for Mann–Whitney U-test, and partial n for ANCOVA. d Missing data for employment/occupation at discharge in 26 cases (19.4%). e ANCOVA for CGI-S and GAF adjusting for time in treatment and the respective baseline scores, means represent adjusted scores.
4. Discussion The present study is the first to assess multidimensional clinical differences between patients with SAD and BD in a representative cohort of patients with a first episode of psychotic mania (FEPM). In this phase of a manic–psychotic syndrome, diagnostic and, subsequently, treatment uncertainties are common. Therefore, longitudinal studies taking into account diagnostic shifts are mandatory (Bromet et al., 2011). Importantly, the design of this study allowed the inclusion of patients with high severity of both psychotic and manic syndromes, low functioning level, lack of insight, non-adherence with medication and high rates of comorbidities (e.g., substance use disorders), many of whom would either have refused or been too ill to participate in a clinical study based on an informed consent procedure (Friis et al., 2003; Menezes et al., 2006). 4.1. Key findings Similar to previous studies (Conus et al., 2010a), we confirmed in a larger and more representative cohort, that patients with an 18-month diagnosis of SAD have lower functioning levels and are less likely to achieve remission of psychotic symptoms compared to those with BD. Further, to our knowledge, this is the first study in FEP, which found a significantly higher medication non-adherence rate in SAD (72%) compared to BD (51%). This finding may partly explain the worse clinical outcome of SAD. The finding that patients with SAD had a nominally lower rate of service disengagement and thus remained in treatment at EPPIC for a longer time is in agreement with EPPIC's experience that patients with higher illness severity during treatment are less likely to disengage from service (Schimmelmann et al., 2006; Conus et al., 2010c). Overall, our results show that among patients with a first episode of psychotic mania, it is difficult to differentiate those who will later be diagnosed with SAD from those who will have a BD diagnosis. In line with Conus et al. (2010a) we found that later SAD diagnosis is associated with a longer DUP. One potential explanation is that a longer DUP increases the likelihood to retrospectively detect psychotic phases without concurrent manic/depressive episodes, which are pathognomonic for SAD. However, a longer DUP was also the best predictor of a diagnostic shift from initial BD to later SAD. This finding indicates that either a longer DUP predicts psychotic phases without concurrent manic/ depressive episodes prospectively throughout the follow-up period or, alternatively, that the retrospective diagnostic assessment of these
pathognomonic psychotic phases is only possible when the patient is sufficiently stabilized and trusting to give valid retrospective information. Another explanation for these findings would be that the late detection and treatment of a psychotic syndrome indeed cause worse multidimensional outcome and thereby the development of SAD rather than BD. It may be worth addressing in quasi-experimental studies comparing services with and without early detection strategies. 4.2. Limitations Medical file audit is a widely used methodology in psychiatry (Olfson et al., 2010) but has been criticized because of the validity of medical records and the lack of inter-rater reliability. We used a range of strategies to minimize such issues as highlighted in the Materials and methods section. The key benefit of such audits is that selection bias linked to informed consent procedures is minimized and a highly representative sample of patients can be obtained (Friis et al., 2003; Menezes et al., 2006). A further limitation of this study was the absence of more standardized measures of psychopathology and premorbid adjustment. It should be noted that this study focused on first-episode psychotic mania patients receiving an 18-month diagnosis of SAD or BD. Thereby, this study is not representative for patients Table 3 Retrospective diagnostic stability in patients with schizoaffective and bipolar I disorder. Final diagnosis Schizoaffective (n = 36; 26.9% ) Baseline diagnosis, n (%) Schizoaffective disorder Bipolar I disorder Schizophreniform disorder Other
Total Bipolar I (n = 98; 73.1%)
(N = 134)
Statistics a
p-value Effect a sizeb b0.001
17 (47.2%)
0
12 (33.3%)
102 (76.1%)
6 (16.7%)
90 (91.8%) 5 (5.1%)
1 (2.8%)
3 (3.1%)
15 (11.2%)
0.68
17 (12.7%)
11 (8.2%)
Bold letters indicate statistically significant differences between SAD and BD (p b 0.05). a Statistics: p-values χ2 for categorical variables, Mann–Whitney U-test for time in treatment, ANCOVA for CGI-S and GAF controlling for time in treatment and the respective baseline values. b Effect size = Cramer-V for χ2 tests, Z/√(N) for Mann–Whitney U-test, and partial n for ANCOVA.
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
presenting with non-psychotic bipolar I disorder, bipolar II disorder or BD presenting with a depression. 4.3. Clinical implications While long DUP may be an indicator of SAD diagnosis or, specifically, of a later shift from BD to SAD, these diagnoses remain difficult to differentiate in patients presenting with a first-episode psychotic–manic syndrome. In future studies, a more detailed assessment of life-time risk factors such as obstetric complications, delays in motor and language development, early cognitive deficits and other biological and psychosocial variables, may yield additional meaningful discriminators of SAD and BD. Nevertheless, our and others' data suggest that the diagnostic entity of SAD is only a helpful construct considering it is linked to a worse course of the illness, more persistent symptoms and psychosocial impairment as well as more risk of medication non-adherence. However, despite these differences, both diagnostic groups need careful monitoring of all symptomatic and functional dimensions to provide optimal treatment. Role of funding source Eli Lilly Australia funded the assessment of files on a subgroup of this cohort treated with olanzapine and risperidone. Eli Lilly did not participate in or influence the data collection, data analyses or write-up of the manuscript. Contributors Drs. Lambert, Conus and McGorry designed this large First Episode Outcome Study. Drs. Lambert and Conus collected the data. Dr. Schimmelmann analyzed the data. Drs. Schöttle, Schimmelmann and Lambert interpreted the data and wrote the first draft of this manuscript. All authors contributed to and have approved the final manuscript. Conflict of interest The acquisition of data was in part supported by the Colonial Foundation, National Health & Medical Research Council and Eli Lilly Australia. Drs. Schimmelmann, Conus, Cotton, McGorry, Naber, Karow and Lambert have received research funding from and/ or served as paid speaker for Eli Lilly. Drs. Schöttle and Michel declare that they have no conflicts of interest. Acknowledgments The study was in part supported by Eli Lilly Australia. Philippe Conus was supported by a grant from the Leenaards Foundation, Switzerland.
References American Psychiatric Association, D.I., 1994. American Psychiatric Association and Task Force on DSM IV, Diagnostic and Statistical Manual of Mental Disorders, DSM IV4 ed. APA, Washington, DC. Bromet, E.J., Kotov, R., Fochtmann, L.J., Carlson, G.A., Tanenberg-Karant, M., Ruggero, C., Chang, S.W., 2011. Diagnostic shifts during the decade following first admission for psychosis. Am. J. Psychiatry 168 (11), 1186–1194. Cheniaux, E., Landeira-Fernandez, J., Lessa Telles, L., Lessa, J.L., Dias, A., Duncan, T., Versiani, M., 2008. Does schizoaffective disorder really exist? A systematic review of the studies that compared schizoaffective disorder with schizophrenia or mood disorders. J. Affect. Disord. 106 (3), 209–217. Cheniaux, E., Landeira-Fernandez, J., Versiani, M., 2009. The diagnoses of schizophrenia, schizoaffective disorder, bipolar disorder and unipolar depression: interrater reliability and congruence between DSM-IV and ICD-10. Psychopathology 42 (5), 293–298. Conus, P., Cotton, S., Schimmelmann, B.G., McGorry, P.D., Lambert, M., 2007. The First Episode Psychosis Outcome Study (FEPOS): pre-morbid and baseline characteristics of 786 first-episode psychosis patients treated in EPPIC between 1998 and 2000. Early Interv. Psychiatry 1, 191–200. Conus, P., Cotton, S., Schimmelmann, B.G., McGorry, P.D., Lambert, M., 2009. Pretreatment and outcome correlates of sexual and physical trauma in an epidemiological cohort of first-episode psychosis patients. Schizophr. Bull. 36 (6), 1105–1114. Conus, P., Abdel-Baki, A., Harrigan, S., Lambert, M., McGorry, P.D., Berk, M., 2010a. Premorbid and outcome correlates of first episode mania with psychosis: is a distinction between schizoaffective and bipolar I disorder valid in the early phase of psychotic disorders? J. Affect. Disord. 126 (1–2), 88–95. Conus, P., Cotton, S., Schimmelmann, B.G., Berk, M., Daglas, R., McGorry, P.D., Lambert, M., 2010b. Pretreatment and outcome correlates of past sexual and physical trauma in 118 bipolar I disorder patients with a first episode of psychotic mania. Bipolar Disord. 12 (3), 244–252. Conus, P., Lambert, M., Cotton, S., Bonsack, C., McGorry, P.D., Schimmelmann, B.G., 2010c. Rate and predictors of service disengagement in an epidemiological firstepisode psychosis cohort. Schizophr. Res. 118 (1–3), 256–263.
35
Cotton, S.M., Lambert, M., Schimmelmann, B.G., Foley, D.L., Morley, K.I., McGorry, P.D., Conus, P., 2009. Gender differences in premorbid, entry, treatment, and outcome characteristics in a treated epidemiological sample of 661 patients with first episode psychosis. Schizophr. Res. 114 (1–3), 17–24. Cotton, S.M., Lambert, M., Schimmelmann, B.G., Mackinnon, A., Gleeson, J.F., Berk, M., Hides, L., Chanen, A., McGorry, P.D., Conus, P., 2012. Depressive symptoms in first episode schizophrenia spectrum disorder. Schizophr. Res. 134 (1), 20–26. Dilling, H., Dittmann, V., 1990. Psychiatric diagnosis following the 10th revision of the International Classification of Diseases (ICD-10). Nervenarzt 61 (5), 259–270. Friis, S., Larsen, T.K., Melle, I., Opjordsmoen, S., Johannessen, J.O., Haahr, U., Simonsen, E., Rund, B.R., Vaglum, P., McGlashan, T., 2003. Methodological pitfalls in early detection studies — the NAPE Lecture 2002. Nordic Association for Psychiatric Epidemiology. Acta Psychiatr. Scand. 107 (1), 3–9. Guy, W., 1976. ECDEU Assessment Manual for Psychopharmacology, Revised. National Institute of Mental Health, Rockville. Lambert, M., Conus, P., Lubman, D.I., Wade, D., Yuen, H., Moritz, S., Naber, D., McGorry, P.D., Schimmelmann, B.G., 2005a. The impact of substance use disorders on clinical outcome in 643 patients with first-episode psychosis. Acta Psychiatr. Scand. 112 (2), 141–148. Lambert, M., Conus, P., Schimmelmann, B.G., Eide, P., Ward, J., Yuen, H., Schacht, M., Edwards, J., Naber, D., McGorry, P.D., 2005b. Comparison of olanzapine and risperidone in 367 first-episode patients with non-affective or affective psychosis: results of an open retrospective medical record study. Pharmacopsychiatry 38 (5), 206–213. Lambert, M., Naber, D., Schacht, A., Wagner, T., Hundemer, H.P., Karow, A., Huber, C.G., Suarez, D., Haro, J.M., Novick, D., Dittmann, R.W., Schimmelmann, B.G., 2008. Rates and predictors of remission and recovery during 3 years in 392 never-treated patients with schizophrenia. Acta Psychiatr. Scand. 118 (3), 220–229. Lambert, M., Conus, P., Cotton, S., Robinson, J., McGorry, P.D., Schimmelmann, B.G., 2010. Prevalence, predictors, and consequences of long-term refusal of antipsychotic treatment in first-episode psychosis. J. Clin. Psychopharmacol. 30 (5), 565–572. McGorry, P.D., Copolov, D.L., Singh, B.S., 1990a. Royal Park Multidiagnostic Instrument for Psychosis: part I. Rationale and review. Schizophr. Bull. 16 (3), 501–515. McGorry, P.D., Singh, B.S., Copolov, D.L., Kaplan, I., Dossetor, C.R., van Riel, R.J., 1990b. Royal Park Multidiagnostic Instrument for Psychosis: part II. Development, reliability, and validity. Schizophr. Bull. 16 (3), 517–536. McGorry, P.D., Edwards, J., Mihalopoulos, C., Harrigan, S.M., Jackson, H.J., 1996. EPPIC: an evolving system of early detection and optimal management. Schizophr. Bull. 22 (2), 305–326. Menezes, N.M., Arenovich, T., Zipursky, R.B., 2006. A systematic review of longitudinal outcome studies of first-episode psychosis. Psychol. Med. 36 (10), 1349–1362. Morley, K.I., Cotton, S.M., Conus, P., Lambert, M., Schimmelmann, B.G., McGorry, P.D., Foley, D.L., 2008. Familial psychopathology in the First Episode Psychosis Outcome Study. Aust. N. Z. J. Psychiatry 42 (7), 617–626. Murru, A., Pacchiarotti, I., Nivoli, A.M., Grande, I., Colom, F., Vieta, E., 2011. What we know and what we don't know about the treatment of schizoaffective disorder. Eur. Neuropsychopharmacol. 21 (9), 680–690. Olfson, M., Marcus, S.C., Doshi, J.A., 2010. Continuity of care after inpatient discharge of patients with schizophrenia in the Medicaid program: a retrospective longitudinal cohort analysis. J. Clin. Psychiatry 71 (7), 831–838. Robinson, D.G., Woerner, M.G., Alvir, J.M., Bilder, R.M., Hinrichsen, G.A., Lieberman, J.A., 2002. Predictors of medication discontinuation by patients with firstepisode schizophrenia and schizoaffective disorder. Schizophr. Res. 57 (2–3), 209–219. Robinson, J., Cotton, S., Conus, P., Schimmelmann, B.G., McGorry, P., Lambert, M., 2009. Prevalence and predictors of suicide attempt in an incidence cohort of 661 young people with first-episode psychosis. Aust. N. Z. J. Psychiatry 43 (2), 149–157. Robinson, J., Harris, M., Cotton, S., Hughes, A., Conus, P., Lambert, M., Schimmelmann, B.G., McGorry, P., 2010. Sudden death among young people with first-episode psychosis: an 8–10 year follow-up study. Psychiatry Res. 177 (3), 305–308. Schimmelmann, B.G., Conus, P., Edwards, J., McGorry, P.D., Lambert, M., 2005. Diagnostic stability 18 months after treatment initiation for first-episode psychosis. J. Clin. Psychiatry 66 (10), 1239–1246. Schimmelmann, B.G., Conus, P., Schacht, M., McGorry, P., Lambert, M., 2006. Predictors of service disengagement in first-admitted adolescents with psychosis. J. Am. Acad. Child Adolesc. Psychiatry 45 (8), 990–999. Schimmelmann, B.G., Conus, P., Cotton, S., McGorry, P.D., Lambert, M., 2007. Pretreatment, baseline, and outcome differences between early-onset and adult-onset psychosis in an epidemiological cohort of 636 first-episode patients. Schizophr. Res. 95 (1–3), 1–8. Schimmelmann, B.G., Huber, C.G., Lambert, M., Cotton, S., McGorry, P.D., Conus, P., 2008. Impact of duration of untreated psychosis on pre-treatment, baseline, and outcome characteristics in an epidemiological first-episode psychosis cohort. J. Psychiatr. Res. 42 (12), 982–990. Schimmelmann, B.G., Conus, P., Cotton, S., Kupferschmid, S., McGorry, P.D., Lambert, M., 2011a. Prevalence and impact of cannabis use disorders in adolescents with early onset first episode psychosis. Eur. Psychiatry. http://dx.doi.org/10.1016/ j.eurpsy.2011.03.001. Schimmelmann, B.G., Conus, P., Cotton, S.M., Kupferschmid, S., Karow, A., Schultze-Lutter, F., McGorry, P.D., Lambert, M., 2011b. Cannabis use disorder and age at onset of psychosis—a study in first-episode patients. Schizophr. Res. 129 (1), 52–56. Spearing, M.K., Post, R.M., Leverich, G.S., Brandt, D., Nolen, W., 1997. Modification of the Clinical Global Impressions (CGI) Scale for use in bipolar illness (BP): the CGI-BP. Psychiatry Res. 73 (3), 159–171.
36
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
Tohen, M., Hennen, J., Zarate Jr., C.M., Baldessarini, R.J., Strakowski, S.M., Stoll, A.L., Faedda, G.L., Suppes, T., Gebre-Medhin, P., Cohen, B.M., 2000a. Two-year syndromal and functional recovery in 219 cases of first-episode major affective disorder with psychotic features. Am. J. Psychiatry 157 (2), 220–228. Tohen, M., Strakowski, S.M., Zarate Jr., C., Hennen, J., Stoll, A.L., Suppes, T., Faedda, G.L., Cohen, B.M., Gebre-Medhin, P., Baldessarini, R.J., 2000b. The McLean–Harvard first-
episode project: 6-month symptomatic and functional outcome in affective and nonaffective psychosis. Biol. Psychiatry 48 (6), 467–476. Ventura, J., Liberman, R.P., Green, M.F., Shaner, A., Mintz, J., 1998. Training and quality assurance with the Structured Clinical Interview for DSM-IV (SCID-I/P). Psychiatry Res. 79 (2), 163–173.
Contents lists available at SciVerse ScienceDirect
Schizophrenia Research journal homepage: www.elsevier.com/locate/schres
Differentiating schizoaffective and bipolar I disorder in first-episode psychotic mania Daniel Schöttle a,⁎, 1, Benno G. Schimmelmann b, 1, Philippe Conus c, Sue M. Cotton d, Chantal Michel b, Patrick D. McGorry d, Anne Karow a, Dieter Naber a, Martin Lambert a a Psychosis Early Detection and Intervention Centre (PEDIC), Centre for Psychosocial Medicine, Department for Psychiatry and Psychotherapy, University Medical Centre Hamburg-Eppendorf, Germany b University Hospital of Child- and Adolescent Psychiatry, University of Bern, Switzerland c Treatment and early Intervention in Psychosis Program (TIPP), Department of Psychiatry CHUV, Lausanne University, Switzerland d Orygen Youth Health Research Centre, Centre for Youth Mental Health, University of Melbourne, Parkville, Australia
a r t i c l e
i n f o
Article history: Received 27 February 2012 Received in revised form 25 June 2012 Accepted 9 July 2012 Available online 28 July 2012 Keywords: Schizoaffective disorder Bipolar I disorder First-episode
a b s t r a c t Objective: This study aims to differentiate schizoaffective disorder (SAD) and bipolar-I-disorder (BD) in first-episode psychotic mania (FEPM). Methods: All 134 patients from an epidemiological first-episode psychosis cohort (N=786) with FEPM and an 18-month follow-up final diagnosis of SAD (n=36) or BD (n=98) were assessed with respect to pre-treatment, baseline and outcome differences. Second, patients with baseline BD who shifted (shifted BD) or did not shift to SAD (stable BD) over the follow-up period were compared regarding pre-treatment and baseline differences. Results: SAD patients displayed a significantly longer duration of untreated psychosis (DUP; effect size r=0.35), a higher illness-severity at baseline (r=0.20) and more traumatic events (Cramer-V=0.19). SAD patients displayed a significantly higher non-adherence rate (Cramer-V=0.19); controlling for time in treatment and respective baseline scores, SAD patients had significantly worse illness severity (CGI-S; partial η2 =0.12) and psychosocial functioning (GAF; partial η2 =0.07) at 18-months, while BD patients were more likely to achieve remission of positive symptoms (OR=4.9, 95% CI=1.8–13.3; p=0.002) and to be employed/occupied (OR=7.7, 95% CI= 2.4–24.4, p=0.001). The main discriminator of stable and shifted BD was a longer DUP in patients shifting from BD to SAD. Conclusions: It is difficult to distinguish BD with psychotic symptoms and SAD in patients presenting with FEPM. Longer DUP is related to SAD and to a shift from BD to SAD. Compared to BD, SAD had worse outcomes and higher rates of non-adherence with medication. Despite these differences, both diagnostic groups need careful dimensional assessment and monitoring of symptoms and functioning in order to choose the right treatment. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Despite the ongoing controversy on the diagnostic entity of schizoaffective disorder (SAD) and its distinction from bipolar I disorder (BD) or schizophrenia (Cheniaux et al., 2008, 2009), it is a clinical reality that a subgroup of patients with psychosis suffers from a combination of mood unrelated psychotic episodes and alternating depressive and manic syndromes with psychotic symptoms. Compared to patients with schizophrenia or bipolar I disorder, such a combination may need specific treatment including pharmacotherapy and psychosocial interventions (Murru et al., 2011).
⁎ Corresponding author at: Psychosis Early Detection and Intervention Centre (PEDIC), Centre for Psychosocial Medicine, Department for Psychiatry and Psychotherapy, University Medical Centre Hamburg-Eppendorf, Martinistrasse 52, 20246 Hamburg, Germany. Tel.: +49 152 22816853; fax: +49 40 741055455. E-mail address: [email protected] (D. Schöttle). 1 Denotes equal contribution to the study. 0920-9964/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.schres.2012.07.010
However, in first-episode psychotic mania it is difficult to foresee whether a patient will eventually have SAD or BD or, specifically, if initially diagnosed with BD, whether the patient will shift to SAD during the course of treatment. In the first and only study on this topic so far, Conus et al. (2010a) reported the following differences between SAD and BD: lower premorbid functioning level, longer duration of prodrome and untreated psychosis, higher severity of positive symptoms at first presentation and, at 12 months follow-up, more severe negative symptoms and a poorer functional level. This study, however, has one important limitation: data were not derived from a representative cohort; this suggests that the inclusion of patients presenting with a high level of severity of both psychotic and manic syndromes and high rates of comorbidities (e.g. substance use disorders) may have been limited. 1.1. Aims of the study The aims of this study were to (1) differentiate patients presenting with first-episode psychotic mania (FEPM) with a discharge diagnosis
32
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
of schizoaffective disorder (SAD) and those with a discharge diagnosis of bipolar I disorder (BD) in terms of pre-treatment, baseline and outcome characteristics and (2) assess pre-treatment and baseline differences between FEPM patients with baseline BD who shifted (=shifted BD) or did not shift to SAD (=stable BD) within an 18-month follow-up period. 2. Materials and methods 2.1. Context and sample The Early Psychosis Prevention and Intervention Centre (EPPIC) is a comprehensive program for young people aged 15–29 years experiencing their first treated episode of psychosis. Treatment spans an average of 18 months. The EPPIC catchment area covers the northwestern regions of Melbourne, Australia. There is a lack of other facilities for the target population in these regions and a scarceness of private psychiatrists. Leakage to private facilities outside the catchment is rare. EPPIC therefore ascertains a treated epidemiological sample of FEP patients (McGorry et al., 1996). The First Episode Psychosis Outcome Study (FEPOS) is a file audit study of all 786 first-episode patients who were admitted to EPPIC between 1998 and 2000 (Lambert et al., 2005b; Conus et al., 2007; Schimmelmann et al., 2007, 2008; Cotton et al., 2009; Robinson et al., 2009; Conus et al., 2010a, 2010b; Robinson et al., 2010; Schimmelmann et al., 2011a, 2011b; Cotton et al., 2012). Of the 786 patients admitted, eighty-two (10%) of the patients' files had been transferred to other services and 43 (5%) patients were excluded because they had a non-psychotic diagnosis at discharge. Of the remaining 661 patients, 182 (27.5%) had a final 18-month diagnosis of SAD or BD. Patients initially presenting with a first-episode manic– psychotic syndrome (regardless of initial psychosis diagnosis) were selected. To ensure the inclusion of clearly manic patients, only those who fulfilled the criterion of at least ≥4 points (moderately to very severely manic) in the mania subscore of the Clinical Global Impression – Scale for use in BD (CGI-BP; Spearing et al., 1997) were included. Data of 134 patients (73.6%) were analyzed, 98 (73%) of those received a final diagnosis of BD and 36 (27%) with a final diagnosis of SAD. 2.2. Assessments and measures For each patient treated at EPPIC, information on pre-treatment, baseline (admission to EPPIC), treatment and outcome characteristics are systematically documented in a structured file. Assessments are based on the Royal Park Multi-diagnostic Instrument for Psychosis (RP-MIP; McGorry et al., 1990a, 1990b). Each file contains information compiled during the 18 month treatment period from various sources using high quality assessments carried out by trained clinicians. Two experienced psychiatrists assessed all files using a standardized questionnaire (Early Psychosis File Questionnaire, EPFQ; Conus et al., 2007). Pre-treatment-, baseline-, treatment and outcome-variable: Premorbid functioning, was assessed with the GAF (American Psychiatric Association, 1994). Baseline-, treatment and outcome-variables: Severity of illness with the CGI-S (Guy, 1976); employment/occupation (yes/no) with the MVCI (Tohen et al., 2000a, 2000b); lack of insight into illness on the basis of one item with anchors ranging from lack of insight to partial and full insight (Conus et al., 2007). Pre-treatment variables: gender; family history of psychosis (yes/no) rated as present in a first or second degree relative (Morley et al., 2008); traumatic history (yes/no) refers to sexual abuse (sexual molestation and/or rape) and/or physical abuse (physical attack or assault or being repeatedly beaten by parents, relatives, or caregivers during childhood (Conus et al., 2009, 2010b)); criminal justice history (yes/no) was considered if there was mention in the file of previous contact with the legal system for any form of conviction; age at onset of psychosis was defined as the age when one first sustained positive psychotic
symptoms occurred according to the DUP scale (McGorry et al., 1990a, 1990b); duration of untreated psychosis with the DUP scale, for detailed procedure see Schimmelmann et al. (2008); lifetime SUD with the DAAS (Lambert et al., 2005a); for detailed procedure see Lambert et al. (2005a); past suicide attempts according to ICD-10 classification (Dilling and Dittmann, 1990). Baseline variables: severity of mania with the CGI-BP (Spearing et al., 1997); age; living with family (yes/no) with the MVCI (Tohen et al., 2000a, 2000b); and co-morbid SUD with the DAAS (Lambert et al., 2005a). Treatment and outcome variables: Time in treatment was calculated on the basis of the regular 18-month treatment in EPPIC and, in the case of service disengagement, the time point when a patient dropped out of treatment; service disengagement was defined as present if patients actively refused any contact with the treatment facility or were not traceable (for detailed procedure see Conus et al., 2010c and Schimmelmann et al., 2006); persistent SUD with the DAAS (for detailed procedure see Lambert et al., 2005a); medication non-adherence was defined according to Robinson et al. (2002) as failure to take medication for 1 week or longer (for detailed procedure see Lambert et al., 2010); suicide attempts in treatment according to ICD-10 classification (Dilling and Dittmann, 1990); hospital admission (yes/no); remission of positive symptoms (yes/no) was defined as receiving a score of no worse than “mild” (score≤ 3) in the CGI-S on discharge from the service or time of service disengagement (according to Lambert et al., 2008). 2.3. Diagnostic assessment, validity, and inter-rater reliability Clinical diagnoses according to DSM-IV criteria (American Psychiatric Association, 1994) at EPPIC are the consensus result of an intensive diagnostic and treatment process, first within the initial 6 weeks of admission by well-trained clinicians working in a specialized assessment and crisis assertive community treatment team, and then at discharge based on all available information. The discharge diagnoses were used to differentiate SAD and BD in this study. Diagnoses were extracted from the files by the principal investigators (ML and PC). In case of disagreement with clinical diagnoses reported in the file, a consensus rating between both research psychiatrists and the patient's case manager was performed (Schimmelmann et al., 2005). Validity of the FEPOS diagnoses was established by the following procedure: between 1998 and 2000, 230 of the 786 patients treated at EPPIC have been included in prospective trials. Their main and co-morbid diagnoses were assessed within 6 weeks of admission using the Structured Clinical Interview for DSM-IV (SCID-I/P; Ventura et al., 1998). The randomly selected SCID and FEPOS diagnoses of 115 patients were compared. The calculated kappa values revealed a very good concordance for both psychosis diagnoses (kappa=0.80) and co-morbid substance use disorder (SUD) diagnoses (kappa=0.74). Inter-rater reliability has been established by comparing baseline ratings given independently by both main investigators in a randomly selected sample of 40 files stratified by time, on the following scales: CGI, GAF, and insight. Analysis revealed a good to very good inter-rater reliability with kappa values ranging from 0.80 to 0.90 (CGI-S: 0.87, GAF score: 0.88, insight score: 0.89). 2.4. Data analysis Patients with discharge SAD and BD were compared with respect to pre-treatment, baseline, and treatment characteristics using Mann– Whitney U-tests when the dependent variable was continuous and chi square analysis (χ2), when the dependent variable was categorical. With respect to outcome differences in terms of CGI-S and GAF, two one-way analyses of covariance (ANCOVAs) were specified controlling for time and treatment and the respective baseline scores. For differences in remission of positive symptoms and employment/occupation rates, two logistic regression analyses were specified with time in service and the respective baseline values (CGI-S and employment/occupation)
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
were entered into the first step, and the final diagnosis of SAD or BD were entered into the second step. For these analyses, the odds ratios (OR) and the 95% confidence interval (CI) of the ORs were calculated and the Wald statistic (z) was used to determine the significance of the associations (only p-values are reported for these analyses). To estimate effect sizes, Cramer-V was used for χ2 tests, the formula r= Z/√(N) for Mann–Whitney U-tests, and partial η2 for ANCOVAs. For χ2 statistics post-hoc examination of standardized residuals (z) was applied to determine which cells contributed to the significant χ2 if appropriate (only p-values will be reported). Pre-treatment and baseline variables of shifted and stable BD were compared applying the above mentioned statistical techniques. The retrospective consistency of the two diagnostic categories was calculated, (i.e., the proportion of patients in a discharge diagnostic category who received the same diagnosis at baseline). All analyses were performed using the IBM® Statistical Package for Social Sciences version 17 (IBM® SPSS® 17.0; Chicago, IL). 3. Results The sample characteristics and differences between patients with a final diagnosis of schizoaffective (SAD) and bipolar I disorder (BD) are displayed in Tables 1 and 2. 3.1. Pre-treatment and baseline differences Patients with SAD had a higher overall rate of traumatic events (44% vs. 25%; p = 0.025), which was due to a higher rate of physical abuse (41% vs. 20%; p = 0.022). Compared to BD, SAD patients were found to have a longer overall DUP (mdn. 31 days vs. 0 days, p b 0.001); specifically the rate of a very short DUP (1–30 days; 42% vs. 69%) was lower in SAD and a long DUP (91–365 days; 31% vs. 4%) was higher in SAD. At baseline, SAD patients had a higher illness severity (CGI-S 6.1 vs. 5.8, p = 0.022). However, the effect sizes for pre-treatment and baseline differences were low except regarding DUP (effect size= 0.35). 3.2. Treatment and outcome differences Patients with SAD stayed in treatment for a significantly longer time (73.0 vs. 59.1 weeks; p =0.041) and were more likely to be nonadherent with medication (72% vs. 51%; p =0.030; see Table 2). No differences were detected regarding service disengagement, persistent substance use, suicide attempts in treatment and rate of patients being hospitalized. Controlling for time in treatment and the respective baseline scores, SAD patients had higher illness severity and worse psychosocial functioning at discharge (adjusted means= 3.2 vs. 2.2; pb 0.001 and 61.5 vs. 69.8; p= 0.002, respectively). The effect sizes for these outcome differences were small. However, controlling for time in treatment and baseline CGI-S, BD patients were clearly more likely to achieve remission of positive symptoms at discharge (OR= 4.9, 95% CI= 1.8–13.3; p = 0.020) and to be employed/occupied (OR= 7.7, 95% CI =2.4–24.4, p= 0.001) compared to SAD patients. 3.3. Pre-treatment and baseline differences between stable and shifting BD In this FEPM sample, the retrospective diagnostic stability of SAD was lower than that of BD (41% vs. 92%; p b 0.001; effect size= 0.68; see Table 3). 33% of patients with a final diagnosis of SAD were initially diagnosed with BD. We further explored whether patients initially diagnosed with BD, who subsequently shifted to SAD (n= 12; shifting BD), differed with regard to pre-treatment and baseline variables from those, who consistently received a BD diagnosis at baseline and discharge (n= 90, stable BD). In other words, it is possible to predict a diagnostic shift to SAD in those, who initially present with BD with psychotic symptoms. As the sample size is low, effect size was chosen as the main statistical
33
Table 1 Pre-treatment and baseline differences between patients with schizoaffective and bipolar disorder. Diagnosis Schizoaffective (n = 36; 26.9% ) Pre-treatment characteristics Male, n (%) 23 (63.9%) Premorbid GAF 71.4 (9.6) score, M (SD) 9 (25.0%) Family history of psychosis, n (%) Traumatic 16 (44.4%) history, n (%) Sexual abuse 8 (22.2%) Physical 13 (40.6%) abuse Criminal justice 9 (26.5%) history, n (%) Age at onset of 22.6 (3.7) psychosis, M (SD) 31 (28–123) DUP, Mdn (quartiles), days DUP-categories 1–30 days 15 (41.7%) 31–90 days 8 (22.2%) 91–365 days 11 (30.6%) >365 days 2 (5.6%) Lifetime SUD, 31 (86.1%) n (%) 3 (8.3%) Past suicide attempts, n (%) Baseline characteristics CGI-S score, 6.1 (0.6) M (SD) CGI-BP mania, 5.1 (0.9) M (SD) GAF score, 26.3 (7.7) M (SD) Age at baseline, 23.0 (3.6) M (SD) in years 19 (52.8%) Employment/ occupation, n (%) 23 (63.9%) Living with family, n (%) Lack of insight, 28 (77.8%) n (%) Co-morbid 27 (75.0%) SUD, n (%)
Total
Statistics
Bipolar I (n = 98; 73.1%)
(N = 134)a
p-value Effect b sizec
58 (59.2%) 74.4 (10.8)
81 (60.4%) 73.6 (10.5)
0.621 0.094
0.04 0.14
26 (28.3%)
35 (27.3%)
0.710
0.03
24 (24.5%)
40 (29.9%)
0.025
0.19
16 (16.3%) 16 (19.8%)
24 (17.9%) 29 (25.7%)
0.430 0.022
0.07 0.22
19 (19.6%)
28 (21.4%)
0.400
0.07
0.462
0.06
b0.001
0.35
22.1 (3.2)
0 (0–31)
68 23 4 3 73
(69.4%) (23.5%) (4.1%) (3.1%) (74.5%)
4 (4.1%)
22.3 (3.3)
30 (0–31)
b0.001
0.39
83 (61.9%) 31 (23.1%) 15 (11.2%) 5 (3.7%) 104 (77.6%)
0.153
0.12
7 (5.3%)
0.334
0.08
5.8 (0.7)
5.9 (0.7)
0.022
0.20
5.3 (0.9)
5.2 (0.9)
0.226
0.10
30.0 (9.2)
29.0 (8.9)
0.094
0.18
22.4 (3.1)
22.5 (3.2)
0.360
0.08
67 (68.4%)
86 (64.2%)
0.095
0.14
48 (49.5%)
71 (53.4%)
0.139
0.13
76 (78.4%)
104 (78.2%)
0.943
0.01
61 (62.2%)
88 (65.7%)
0.168
0.12
Bold letters indicate statistically significant differences between SAD and BD (p b 0.05). Abbreviations: GAF = Global Assessment of Functioning Scale, DUP = duration of untreated psychosis, SUD = substance use disorder, CGI = Clinical Global Impression — Severity of Illness Scale, M = mean, Mdn= median, SD = standard deviation. a Ns vary due to missing data. b Descriptive statistics: p-values χ2 for categorical variables, Mann–Whitney U-tests for continuous pre-treatment and baseline characteristics. c Effect size = Cramer-V for χ2 tests and Z/√(N) for Mann–Whitney U-tests.
descriptor. Only differences with an effect size at least approaching moderate level, i.e. of at least 0.25, were considered indicative of a noteworthy group effect. Only DUP-categories were different between stable and shifting BD: the rate of a very short DUP (1–30 days; 50% vs. 69%) was lower in stable BD and a long DUP (91–365 days; 33% vs. 4%) was higher in shifting BD (Fisher-test; χ2 =12.4; p =0.022; effect size= 0.35).
34
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
Table 2 Treatment and outcome differences between patients with schizoaffective and bipolar I disorder. Diagnosis
Time in treatment, M (SD) weeks Service disengagement, n (%) Persistent SUD, n (%) Medication non-adherence, n (%) Suicide attempts in treatment, n (%) Hospital admission, n (%) Remission of positive symptoms, n (%) Employment/occupation d, n (%) CGI-S score e, M (SD) GAF score e, M (SD) Lack of insight, n (%)
Total
Statistics
Schizoaffective (n = 36; 26.9% )
Bipolar I (n = 98; 73.1%)
(N = 134)a
p-valueb
Effect sizec
73.0 (27.6) 4 (11.1%) 8 (22.2%) 26 (72.2%) 4 (11.1%) 35 (97.2%) 22 (61.1%) 6 (22.2%) 3.2 (1.3) 61.5 (13.6) 5 (13.9%)
59.1 (33.7) 20 (20.4%) 19 (19.4%) 47 (51.1%) 6 (6.1%) 88 (89.8%) 79 (80.6%) 50 (60.2%) 2.2 (1.2) 69.8 (14.2) 6 (6.1%)
62.9 24 27 73 10 123 101 56 2.5 67.4 11
0.041 0.213 0.717 0.030 0.330 0.165 0.020 0.001 b0.001 0.002 0.147
0.18 0.11 0.03 0.19 0.08 0.12 0.20 0.33 0.12 0.07 0.13
(32.7) (17.9%) (20.1%) (57.0%) (7.5%) (91.8%) (75.4%) (50.9%) (1.3) (14.4) (8.2%)
Bold letters indicate statistically significant differences between SAD and BD (p b 0.05). Abbreviations: GAF = Global Assessment of Functioning Scale, DUP = duration of untreated psychosis, SUD = substance use disorder, CGI = Clinical Global Impression — Severity of Illness Scale, M = mean, Mdn = median, SD = standard deviation. a Ns vary due to missing data. b Statistics: p-values χ2 for categorical variables, Mann–Whitney U-test for time in treatment, ANCOVA for CGI-S and GAF controlling for time in treatment and the respective baseline values. c Effect size = Cramer-V for χ2 tests, Z/√(N) for Mann–Whitney U-test, and partial n for ANCOVA. d Missing data for employment/occupation at discharge in 26 cases (19.4%). e ANCOVA for CGI-S and GAF adjusting for time in treatment and the respective baseline scores, means represent adjusted scores.
4. Discussion The present study is the first to assess multidimensional clinical differences between patients with SAD and BD in a representative cohort of patients with a first episode of psychotic mania (FEPM). In this phase of a manic–psychotic syndrome, diagnostic and, subsequently, treatment uncertainties are common. Therefore, longitudinal studies taking into account diagnostic shifts are mandatory (Bromet et al., 2011). Importantly, the design of this study allowed the inclusion of patients with high severity of both psychotic and manic syndromes, low functioning level, lack of insight, non-adherence with medication and high rates of comorbidities (e.g., substance use disorders), many of whom would either have refused or been too ill to participate in a clinical study based on an informed consent procedure (Friis et al., 2003; Menezes et al., 2006). 4.1. Key findings Similar to previous studies (Conus et al., 2010a), we confirmed in a larger and more representative cohort, that patients with an 18-month diagnosis of SAD have lower functioning levels and are less likely to achieve remission of psychotic symptoms compared to those with BD. Further, to our knowledge, this is the first study in FEP, which found a significantly higher medication non-adherence rate in SAD (72%) compared to BD (51%). This finding may partly explain the worse clinical outcome of SAD. The finding that patients with SAD had a nominally lower rate of service disengagement and thus remained in treatment at EPPIC for a longer time is in agreement with EPPIC's experience that patients with higher illness severity during treatment are less likely to disengage from service (Schimmelmann et al., 2006; Conus et al., 2010c). Overall, our results show that among patients with a first episode of psychotic mania, it is difficult to differentiate those who will later be diagnosed with SAD from those who will have a BD diagnosis. In line with Conus et al. (2010a) we found that later SAD diagnosis is associated with a longer DUP. One potential explanation is that a longer DUP increases the likelihood to retrospectively detect psychotic phases without concurrent manic/depressive episodes, which are pathognomonic for SAD. However, a longer DUP was also the best predictor of a diagnostic shift from initial BD to later SAD. This finding indicates that either a longer DUP predicts psychotic phases without concurrent manic/ depressive episodes prospectively throughout the follow-up period or, alternatively, that the retrospective diagnostic assessment of these
pathognomonic psychotic phases is only possible when the patient is sufficiently stabilized and trusting to give valid retrospective information. Another explanation for these findings would be that the late detection and treatment of a psychotic syndrome indeed cause worse multidimensional outcome and thereby the development of SAD rather than BD. It may be worth addressing in quasi-experimental studies comparing services with and without early detection strategies. 4.2. Limitations Medical file audit is a widely used methodology in psychiatry (Olfson et al., 2010) but has been criticized because of the validity of medical records and the lack of inter-rater reliability. We used a range of strategies to minimize such issues as highlighted in the Materials and methods section. The key benefit of such audits is that selection bias linked to informed consent procedures is minimized and a highly representative sample of patients can be obtained (Friis et al., 2003; Menezes et al., 2006). A further limitation of this study was the absence of more standardized measures of psychopathology and premorbid adjustment. It should be noted that this study focused on first-episode psychotic mania patients receiving an 18-month diagnosis of SAD or BD. Thereby, this study is not representative for patients Table 3 Retrospective diagnostic stability in patients with schizoaffective and bipolar I disorder. Final diagnosis Schizoaffective (n = 36; 26.9% ) Baseline diagnosis, n (%) Schizoaffective disorder Bipolar I disorder Schizophreniform disorder Other
Total Bipolar I (n = 98; 73.1%)
(N = 134)
Statistics a
p-value Effect a sizeb b0.001
17 (47.2%)
0
12 (33.3%)
102 (76.1%)
6 (16.7%)
90 (91.8%) 5 (5.1%)
1 (2.8%)
3 (3.1%)
15 (11.2%)
0.68
17 (12.7%)
11 (8.2%)
Bold letters indicate statistically significant differences between SAD and BD (p b 0.05). a Statistics: p-values χ2 for categorical variables, Mann–Whitney U-test for time in treatment, ANCOVA for CGI-S and GAF controlling for time in treatment and the respective baseline values. b Effect size = Cramer-V for χ2 tests, Z/√(N) for Mann–Whitney U-test, and partial n for ANCOVA.
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
presenting with non-psychotic bipolar I disorder, bipolar II disorder or BD presenting with a depression. 4.3. Clinical implications While long DUP may be an indicator of SAD diagnosis or, specifically, of a later shift from BD to SAD, these diagnoses remain difficult to differentiate in patients presenting with a first-episode psychotic–manic syndrome. In future studies, a more detailed assessment of life-time risk factors such as obstetric complications, delays in motor and language development, early cognitive deficits and other biological and psychosocial variables, may yield additional meaningful discriminators of SAD and BD. Nevertheless, our and others' data suggest that the diagnostic entity of SAD is only a helpful construct considering it is linked to a worse course of the illness, more persistent symptoms and psychosocial impairment as well as more risk of medication non-adherence. However, despite these differences, both diagnostic groups need careful monitoring of all symptomatic and functional dimensions to provide optimal treatment. Role of funding source Eli Lilly Australia funded the assessment of files on a subgroup of this cohort treated with olanzapine and risperidone. Eli Lilly did not participate in or influence the data collection, data analyses or write-up of the manuscript. Contributors Drs. Lambert, Conus and McGorry designed this large First Episode Outcome Study. Drs. Lambert and Conus collected the data. Dr. Schimmelmann analyzed the data. Drs. Schöttle, Schimmelmann and Lambert interpreted the data and wrote the first draft of this manuscript. All authors contributed to and have approved the final manuscript. Conflict of interest The acquisition of data was in part supported by the Colonial Foundation, National Health & Medical Research Council and Eli Lilly Australia. Drs. Schimmelmann, Conus, Cotton, McGorry, Naber, Karow and Lambert have received research funding from and/ or served as paid speaker for Eli Lilly. Drs. Schöttle and Michel declare that they have no conflicts of interest. Acknowledgments The study was in part supported by Eli Lilly Australia. Philippe Conus was supported by a grant from the Leenaards Foundation, Switzerland.
References American Psychiatric Association, D.I., 1994. American Psychiatric Association and Task Force on DSM IV, Diagnostic and Statistical Manual of Mental Disorders, DSM IV4 ed. APA, Washington, DC. Bromet, E.J., Kotov, R., Fochtmann, L.J., Carlson, G.A., Tanenberg-Karant, M., Ruggero, C., Chang, S.W., 2011. Diagnostic shifts during the decade following first admission for psychosis. Am. J. Psychiatry 168 (11), 1186–1194. Cheniaux, E., Landeira-Fernandez, J., Lessa Telles, L., Lessa, J.L., Dias, A., Duncan, T., Versiani, M., 2008. Does schizoaffective disorder really exist? A systematic review of the studies that compared schizoaffective disorder with schizophrenia or mood disorders. J. Affect. Disord. 106 (3), 209–217. Cheniaux, E., Landeira-Fernandez, J., Versiani, M., 2009. The diagnoses of schizophrenia, schizoaffective disorder, bipolar disorder and unipolar depression: interrater reliability and congruence between DSM-IV and ICD-10. Psychopathology 42 (5), 293–298. Conus, P., Cotton, S., Schimmelmann, B.G., McGorry, P.D., Lambert, M., 2007. The First Episode Psychosis Outcome Study (FEPOS): pre-morbid and baseline characteristics of 786 first-episode psychosis patients treated in EPPIC between 1998 and 2000. Early Interv. Psychiatry 1, 191–200. Conus, P., Cotton, S., Schimmelmann, B.G., McGorry, P.D., Lambert, M., 2009. Pretreatment and outcome correlates of sexual and physical trauma in an epidemiological cohort of first-episode psychosis patients. Schizophr. Bull. 36 (6), 1105–1114. Conus, P., Abdel-Baki, A., Harrigan, S., Lambert, M., McGorry, P.D., Berk, M., 2010a. Premorbid and outcome correlates of first episode mania with psychosis: is a distinction between schizoaffective and bipolar I disorder valid in the early phase of psychotic disorders? J. Affect. Disord. 126 (1–2), 88–95. Conus, P., Cotton, S., Schimmelmann, B.G., Berk, M., Daglas, R., McGorry, P.D., Lambert, M., 2010b. Pretreatment and outcome correlates of past sexual and physical trauma in 118 bipolar I disorder patients with a first episode of psychotic mania. Bipolar Disord. 12 (3), 244–252. Conus, P., Lambert, M., Cotton, S., Bonsack, C., McGorry, P.D., Schimmelmann, B.G., 2010c. Rate and predictors of service disengagement in an epidemiological firstepisode psychosis cohort. Schizophr. Res. 118 (1–3), 256–263.
35
Cotton, S.M., Lambert, M., Schimmelmann, B.G., Foley, D.L., Morley, K.I., McGorry, P.D., Conus, P., 2009. Gender differences in premorbid, entry, treatment, and outcome characteristics in a treated epidemiological sample of 661 patients with first episode psychosis. Schizophr. Res. 114 (1–3), 17–24. Cotton, S.M., Lambert, M., Schimmelmann, B.G., Mackinnon, A., Gleeson, J.F., Berk, M., Hides, L., Chanen, A., McGorry, P.D., Conus, P., 2012. Depressive symptoms in first episode schizophrenia spectrum disorder. Schizophr. Res. 134 (1), 20–26. Dilling, H., Dittmann, V., 1990. Psychiatric diagnosis following the 10th revision of the International Classification of Diseases (ICD-10). Nervenarzt 61 (5), 259–270. Friis, S., Larsen, T.K., Melle, I., Opjordsmoen, S., Johannessen, J.O., Haahr, U., Simonsen, E., Rund, B.R., Vaglum, P., McGlashan, T., 2003. Methodological pitfalls in early detection studies — the NAPE Lecture 2002. Nordic Association for Psychiatric Epidemiology. Acta Psychiatr. Scand. 107 (1), 3–9. Guy, W., 1976. ECDEU Assessment Manual for Psychopharmacology, Revised. National Institute of Mental Health, Rockville. Lambert, M., Conus, P., Lubman, D.I., Wade, D., Yuen, H., Moritz, S., Naber, D., McGorry, P.D., Schimmelmann, B.G., 2005a. The impact of substance use disorders on clinical outcome in 643 patients with first-episode psychosis. Acta Psychiatr. Scand. 112 (2), 141–148. Lambert, M., Conus, P., Schimmelmann, B.G., Eide, P., Ward, J., Yuen, H., Schacht, M., Edwards, J., Naber, D., McGorry, P.D., 2005b. Comparison of olanzapine and risperidone in 367 first-episode patients with non-affective or affective psychosis: results of an open retrospective medical record study. Pharmacopsychiatry 38 (5), 206–213. Lambert, M., Naber, D., Schacht, A., Wagner, T., Hundemer, H.P., Karow, A., Huber, C.G., Suarez, D., Haro, J.M., Novick, D., Dittmann, R.W., Schimmelmann, B.G., 2008. Rates and predictors of remission and recovery during 3 years in 392 never-treated patients with schizophrenia. Acta Psychiatr. Scand. 118 (3), 220–229. Lambert, M., Conus, P., Cotton, S., Robinson, J., McGorry, P.D., Schimmelmann, B.G., 2010. Prevalence, predictors, and consequences of long-term refusal of antipsychotic treatment in first-episode psychosis. J. Clin. Psychopharmacol. 30 (5), 565–572. McGorry, P.D., Copolov, D.L., Singh, B.S., 1990a. Royal Park Multidiagnostic Instrument for Psychosis: part I. Rationale and review. Schizophr. Bull. 16 (3), 501–515. McGorry, P.D., Singh, B.S., Copolov, D.L., Kaplan, I., Dossetor, C.R., van Riel, R.J., 1990b. Royal Park Multidiagnostic Instrument for Psychosis: part II. Development, reliability, and validity. Schizophr. Bull. 16 (3), 517–536. McGorry, P.D., Edwards, J., Mihalopoulos, C., Harrigan, S.M., Jackson, H.J., 1996. EPPIC: an evolving system of early detection and optimal management. Schizophr. Bull. 22 (2), 305–326. Menezes, N.M., Arenovich, T., Zipursky, R.B., 2006. A systematic review of longitudinal outcome studies of first-episode psychosis. Psychol. Med. 36 (10), 1349–1362. Morley, K.I., Cotton, S.M., Conus, P., Lambert, M., Schimmelmann, B.G., McGorry, P.D., Foley, D.L., 2008. Familial psychopathology in the First Episode Psychosis Outcome Study. Aust. N. Z. J. Psychiatry 42 (7), 617–626. Murru, A., Pacchiarotti, I., Nivoli, A.M., Grande, I., Colom, F., Vieta, E., 2011. What we know and what we don't know about the treatment of schizoaffective disorder. Eur. Neuropsychopharmacol. 21 (9), 680–690. Olfson, M., Marcus, S.C., Doshi, J.A., 2010. Continuity of care after inpatient discharge of patients with schizophrenia in the Medicaid program: a retrospective longitudinal cohort analysis. J. Clin. Psychiatry 71 (7), 831–838. Robinson, D.G., Woerner, M.G., Alvir, J.M., Bilder, R.M., Hinrichsen, G.A., Lieberman, J.A., 2002. Predictors of medication discontinuation by patients with firstepisode schizophrenia and schizoaffective disorder. Schizophr. Res. 57 (2–3), 209–219. Robinson, J., Cotton, S., Conus, P., Schimmelmann, B.G., McGorry, P., Lambert, M., 2009. Prevalence and predictors of suicide attempt in an incidence cohort of 661 young people with first-episode psychosis. Aust. N. Z. J. Psychiatry 43 (2), 149–157. Robinson, J., Harris, M., Cotton, S., Hughes, A., Conus, P., Lambert, M., Schimmelmann, B.G., McGorry, P., 2010. Sudden death among young people with first-episode psychosis: an 8–10 year follow-up study. Psychiatry Res. 177 (3), 305–308. Schimmelmann, B.G., Conus, P., Edwards, J., McGorry, P.D., Lambert, M., 2005. Diagnostic stability 18 months after treatment initiation for first-episode psychosis. J. Clin. Psychiatry 66 (10), 1239–1246. Schimmelmann, B.G., Conus, P., Schacht, M., McGorry, P., Lambert, M., 2006. Predictors of service disengagement in first-admitted adolescents with psychosis. J. Am. Acad. Child Adolesc. Psychiatry 45 (8), 990–999. Schimmelmann, B.G., Conus, P., Cotton, S., McGorry, P.D., Lambert, M., 2007. Pretreatment, baseline, and outcome differences between early-onset and adult-onset psychosis in an epidemiological cohort of 636 first-episode patients. Schizophr. Res. 95 (1–3), 1–8. Schimmelmann, B.G., Huber, C.G., Lambert, M., Cotton, S., McGorry, P.D., Conus, P., 2008. Impact of duration of untreated psychosis on pre-treatment, baseline, and outcome characteristics in an epidemiological first-episode psychosis cohort. J. Psychiatr. Res. 42 (12), 982–990. Schimmelmann, B.G., Conus, P., Cotton, S., Kupferschmid, S., McGorry, P.D., Lambert, M., 2011a. Prevalence and impact of cannabis use disorders in adolescents with early onset first episode psychosis. Eur. Psychiatry. http://dx.doi.org/10.1016/ j.eurpsy.2011.03.001. Schimmelmann, B.G., Conus, P., Cotton, S.M., Kupferschmid, S., Karow, A., Schultze-Lutter, F., McGorry, P.D., Lambert, M., 2011b. Cannabis use disorder and age at onset of psychosis—a study in first-episode patients. Schizophr. Res. 129 (1), 52–56. Spearing, M.K., Post, R.M., Leverich, G.S., Brandt, D., Nolen, W., 1997. Modification of the Clinical Global Impressions (CGI) Scale for use in bipolar illness (BP): the CGI-BP. Psychiatry Res. 73 (3), 159–171.
36
D. Schöttle et al. / Schizophrenia Research 140 (2012) 31–36
Tohen, M., Hennen, J., Zarate Jr., C.M., Baldessarini, R.J., Strakowski, S.M., Stoll, A.L., Faedda, G.L., Suppes, T., Gebre-Medhin, P., Cohen, B.M., 2000a. Two-year syndromal and functional recovery in 219 cases of first-episode major affective disorder with psychotic features. Am. J. Psychiatry 157 (2), 220–228. Tohen, M., Strakowski, S.M., Zarate Jr., C., Hennen, J., Stoll, A.L., Suppes, T., Faedda, G.L., Cohen, B.M., Gebre-Medhin, P., Baldessarini, R.J., 2000b. The McLean–Harvard first-
episode project: 6-month symptomatic and functional outcome in affective and nonaffective psychosis. Biol. Psychiatry 48 (6), 467–476. Ventura, J., Liberman, R.P., Green, M.F., Shaner, A., Mintz, J., 1998. Training and quality assurance with the Structured Clinical Interview for DSM-IV (SCID-I/P). Psychiatry Res. 79 (2), 163–173.