Eur Psychiatry (2015) 30a

European Psychiatry 30 (2015) 405–416
Contents lists available at ScienceDirect
European Psychiatry
journal homepage: http://www.europsy-journal.com
Original article
EPA guidance on the early detection of clinical high risk states of

psychoses
F. Schultze-Lutter a, C. Michel a, S.J. Schmidt a, B.G. Schimmelmann a, N.P. Maric c,
R.K.R. Salokangas d, A. Riecher-Rössler e, M. van der Gaag f,g, M. Nordentoft h, A. Raballo i,j,
A. Meneghelli k, M. Marshall l,m, A. Morrison n,o, S. Ruhrmann b, J. Klosterkötter b,*
a
University Hospital of Child and Adolescent Psychiatry and Psychotherapy, University of Bern, Bern, Switzerland
b
Department of Psychiatry and Psychotherapy, University of Cologne, Cologne, Germany
c
School of Medicine, University of Belgrade and Clinic of Psychiatry, Clinical Center of Serbia, Belgrade, Serbia
d
Department of Psychiatry, University of Turku, Turku, Finland
e
Center for Gender Research and Early Detection, Psychiatric University Clinics Basel, Basel, Switzerland
f
Department of Clinical Psychology, VU University and EMGO Institute for Health and Care Research, Amsterdam, The Netherlands
g
Psychosis Research, Parnassia Psychiatric Institute, The Hague, The Netherlands
h
Mental Health Center Copenhagen, University of Copenhagen, Copenhagen, Denmark
i
Department of Mental Health, Reggio Emilia Public Health Centre, Reggio Emilia, Italy
j
Regional Working Group on Early Detection of Psychosis, Emilia Romagna Regional Health Service, Bologna, Italy
k
Dipartimento di Salute Mentale, Centro per l’Individuazione e l’Intervento Precoce nelle Psicosi-Programma 2000, Ospedale Niguarda Ca’ Granda, Milan, Italy
l
School of Medicine, University of Manchester, Manchester, UK
m
LANTERN Centre, Lancashire Care NHS Foundation Trust, Preston, UK
n
School of Psychological Sciences, University of Manchester, Manchester, UK
o
Psychosis Research Unit, Greater Manchester West NHS Mental Health Trust, Manchester, UK
A R T I C L E I N F O A B S T R A C T
Article history: The aim of this guidance paper of the European Psychiatric Association is to provide evidence-based
Received 10 November 2014 recommendations on the early detection of a clinical high risk (CHR) for psychosis in patients with
Received in revised form 29 January 2015 mental problems. To this aim, we conducted a meta-analysis of studies reporting on conversion rates to
Accepted 29 January 2015
psychosis in non-overlapping samples meeting any at least any one of the main CHR criteria: ultra-high
Available online 27 February 2015
risk (UHR) and/or basic symptoms criteria. Further, effects of potential moderators (different UHR
criteria definitions, single UHR criteria and age) on conversion rates were examined. Conversion rates in
Keywords:
the identified 42 samples with altogether more than 4000 CHR patients who had mainly been identified
Meta-analysis
Prevention of psychosis in Europe
by UHR criteria and/or the basic symptom criterion ‘cognitive disturbances’ (COGDIS) showed
Attenuated psychotic symptoms (APS) considerable heterogeneity. While UHR criteria and COGDIS were related to similar conversion rates
Transient psychotic symptoms (BLIPS) until 2-year follow-up, conversion rates of COGDIS were significantly higher thereafter. Differences in
Cognitive disturbances (COGDIS) onset and frequency requirements of symptomatic UHR criteria or in their different consideration of
Children and adolescents functional decline, substance use and co-morbidity did not seem to impact on conversion rates. The
‘genetic risk and functional decline’ UHR criterion was rarely met and only showed an insignificant
pooled sample effect. However, age significantly affected UHR conversion rates with lower rates in
children and adolescents. Although more research into potential sources of heterogeneity in conversion
rates is needed to facilitate improvement of CHR criteria, six evidence-based recommendations for an
early detection of psychosis were developed as a basis for the EPA guidance on early intervention in CHR
states.
ß 2015 Elsevier Masson SAS. All rights reserved.
1. Introduction
In psychiatry, as in medicine, strenuous efforts are made to

* Corresponding author at: Department of Psychiatry and Psychotherapy,
predict and, subsequently, prevent diseases before their first
University of Cologne, Kerpener Strasse 62, 50924 Cologne, Germany.
Tel.: +49 221 478 4001; fax: +49 221 478 87139. manifestation and the development of significant disability
E-mail address: [email protected] (J. Klosterkötter). [13,51,127]. In psychosis research, this approach has already been
http://dx.doi.org/10.1016/j.eurpsy.2015.01.010
0924-9338/ß 2015 Elsevier Masson SAS. All rights reserved.
406 F. Schultze-Lutter et al. / European Psychiatry 30 (2015) 405–416
pursued over the past two decades within the framework of interactions in development of psychoses is conducted intensively
indicated prevention in help-seeking samples [25,127]. Since a in Europe [22].
successful preventive intervention relies on the accuracy of risk
detection, the present paper critically examines present research 1.3. Rationale for a prevention of psychoses
on the detection of clinical high risk (CHR) states to underpin the
development of clinical recommendations that reflect current The epidemiological, clinical and etiopathogenic aspects of
evidence in this sensitive and changing area of research. A second psychoses outlined above, and the lack of a therapeutic break-
related paper (see Schmidt SJ et al.; this issue) will examine the through in the treatment of the disorder itself make psychotic
evidence for preventive interventions in this area and provide disorders a worthwhile target for preventive measures prior to
clinical recommendations. Together, both papers offer up to date their first manifestation. In principle, prevention can be offered:
evidence-based guidance for both the prediction and the preven- universally to the general, unselected population; selectively to
tion of psychosis with special emphasis on potential developmen- healthy individuals with a known risk factor of the disease; or by
tal aspects. indication to persons already suffering from first complaints and
impairments and who are actively seeking advice and help
1.1. Prevalence and burden of psychotic disorders [73,127]. The universal and the selective approach cannot be
implemented effectively–at least to date–due to: the low incidence
The defining characteristic of psychosis is the presence of of psychoses in the general population, lack of sufficient etiological
positive symptoms, i.e. delusions, hallucinations and positive knowledge and of risk factors of sufficiently large effect. The
formal thought disorders, yet again confirmed as the key features indicated approach is currently regarded as the most appropriate
of psychotic disorders in DSM-5 [4]. The lifetime prevalence of prevention strategy for psychoses [51], because the majority of
psychoses is estimated between 0.2 and 3.5% [83,125], their annual first-episode psychosis patients report having suffered from
incidence between 0.01 and 0.035%, with growing numbers mental problems including risk symptoms and increasing psycho-
reported in Europe where, within 12 months, approximately social impairment for an average 5-year period prior to the onset of
3.7 million adults (0.8%) had been affected in 2005 and as much psychosis [106] (Fig. 1). This strategy is supported by consistently
as 5 million (1.2%) in 2011 [46,125]. The gender related incidence reported negative effects of long duration of untreated illness and
of affective and non-affective psychotic disorders depends on type untreated psychosis on outcome [29,61] that may even be
of psychosis and age with a higher incidence of schizophrenia in aggravated in EOP, because more pronounced neurodevelopmen-
men and a similar cumulative incidence of all psychoses at age 60 tal and cognitive deficits, the insidious onset of less pronounced
[19,34,35,38,47,65,114]. Approximately 10–15% of all psychoses positive symptoms and/or the atypical clinical picture of the
are early-onset psychoses (EOP) manifesting before the age of 18, beginning EOP–potentially misinterpreted as ‘adolescent crisis’–
and approximately 1–3% are very-early-onset psychoses (VEOP) might act as further delaying factors [96,97].
with an onset before the age of 13 [98,125].
Following psychotic episodes, negative symptoms commonly 1.4. The clinical high risk (CHR) state of psychoses
persist, and are associated with cognitive impairments and
psychosocial disabilities. This is a main reason why such a Currently, there are two complementary approaches to the
relatively infrequent disorder is responsible for the sixth largest characterization of the CHR state of psychoses: the ultra-high risk
share of disability-adjusted life years (DALYs) in adults in Europe (UHR) and the basic symptoms criteria (Fig. 1) [25,51]. The
(i.e., 637,693 DALYs [125]), and the third largest (16.8 million alternative UHR criteria, which comprise the attenuated psychotic
DALYs) of all main brain disorders worldwide [16]. Despite the symptom (APS) criterion, the brief limited intermittent psychotic
infrequency of (V)EOP, schizophrenia is one of the ten main causes symptom (BLIPS) criterion, and the genetic risk and functional
of DALYs in 10- to 14-year-old boys and 15- to 19-year-old girls decline (GRFD) criterion (Table 1), were originally developed with
[31]. Thus, at s 93.9 billion of total direct health care, direct non- the explicit aim of detecting an imminent risk for psychoses, i.e.,
medical and indirect costs of brain disorders in Europe in persons at risk for developing a first-episode within the next
2010 attributed to psychoses, only the costs for mood disorders 12 months [84]. While their operationalization usually hardly
and dementia were higher [81]. In addition, the burden caused differs with respect to these broad definitions, the associated
by stigma and discrimination is also among the highest in requirements in particular of APS and BLIPS criteria can differ
psychosis [89]. considerably between assessments (Table 2) [110]. Table 3 details
instruments used for the assessment of UHR criteria.
1.2. Etiological and pathogenetic aspects in psychoses In contrast to the UHR criteria, the criteria based on basic
symptoms (the cognitive-perceptive basic symptoms (COPER)
Psychoses are increasingly considered as a brain development criterion and the cognitive disturbances (COGDIS) criterion
disorder with polygenic heredity [36]. As with other complex (Table 4) [50,102,108]) were developed to detect the risk for
diseases, research is now focusing on characterizing the polygenic psychosis as early as possible in the development of the illness,
factors and clarifying their variable phenotypic expression. This ideally before functional impairments appeared (Fig. 1). Basic
pathogenesis seems to be greatly influenced by both rare gene symptoms are currently assessed with the Schizophrenia
variants with large effects, and interactions between different Proneness Instrument, Adult ((SPI-A [104]) or Child & Youth
genes of small effect as well as genes and environment version (SPI-CY [109])).
[118]. Contributory environmental risk factors include exposure
to viral agents in the second trimester of pregnancy, birth 1.5. Early detection of psychoses in children and adolescents
complications, childhood trauma, migration, the quality of the
rearing environment, environment, socio-economic disadvantage, Since EOP were reported to present a slightly different
urban birth, living in urban areas and using illicit drugs, onset and clinical picture compared to adult-onset psychoses
particularly cannabis. However, with odds ratios of around 2, [2,6,18,27,37,88,90,94,115,108], early detection in children and
each of these factors increase lifetime-risk for psychosis only adolescents might be confronted with additional challenges. This is
slightly [117] and causality can be difficult to determine. Thus, to supported by first reports on conversion rates in adolescent risk
improve future prediction, research on gene environment samples between age 12 and 18 [122,135], indicating that lag time
F. Schultze-Lutter et al. / European Psychiatry 30 (2015) 405–416 407
Fig. 1. Model of the early course of psychosis based on Fusar-Poli et al. (2012) [25].
to conversion might be longer and, consequently, conversion rates this paper aims to reflect the current state of evidence of the
in the first years following initial risk assessment might be lower. different CHR criteria in different age groups, and to make
Furthermore, recent studies reported high prevalence rates of evidence-based recommendations for their clinical use in Europe.
(attenuated) psychotic symptoms [111], in particular of hallucina-
tions, in children and young adolescents, which seem to decrease 2. Methods
with age [43,44] and to remit spontaneously in about three
quarters [7]. Thus, it was recently argued that the validity of 2.1. Literature selection
current risk criteria needs to be examined in and possibly adapted
to children and adolescents [23,95,99,107]. 2.1.1. Literature search
We conducted a systematic literature review in June 2014 in
1.6. Aims PubMed and Scopus that covered all journals included in Embase
using the following search terms and syntax: ([early detection] OR
With studies on early detection of psychosis accumulating over [prediction] OR [early recognition]) AND ([conversion] OR [transition]
the past 20 years and growing interest in this field from clinicians, OR [development]) AND ([psychosis] OR [schizophrenia]) AND ([risk]
OR [prodrome]). Since the early detection of psychosis is a
Table 1
predominately psychiatric topic, an additional search in PsycInfo
General definition of ultra-high risk (UHR) criteria. was not conducted as it covers fewer psychiatric journals than
Symptomatic approach
PubMed and Scopus.
A. Presence of at least any 1 brief limited intermittent psychotic symptom
(BLIPS) 2.1.2. Selection criteria
Hallucinations Inclusion criteria were:
Delusions
Formal thought disorders
B. Presence of at least any 1 attenuated psychotic symptom (APS) study was prospective with a (mean) follow-up of at least
Ideas of reference 6 months;
Odd beliefs or magical thinking, including ideas of grandiosity study reported on a CHR sample according to the UHR or basic
Paranoid ideation
symptom criteria;
Unusual perceptual experiences
Odd thinking and speech
primary or secondary outcome was psychosis and;
paper was published in German or English.
Constrictive ‘state-trait’ approach
C. Presence of a genetic risk factor (family history of psychosis; schizotypal
personality disorder of person) in combination with a recent significant
Exclusion criteria were:
decline in psychosocial functioning (GRFD)
study was published before 1996, i.e., before the first description
Presence of at least any one of A, B or C to meet UHR criteria
of main CHR criteria;
Table 2
Comparison of additional requirements of symptomatic ultra-high risk (UHR) criteria in the Structured Interview for Psychosis-Risk Syndromes (SIPS [64]) and the
Comprehensive Assessment of At-Risk Mental States (CAARMS) early versions [130] as well as latest 2006 version [131].
Scale Onset Frequency Substance-use, co-morbidities Functioning
Attenuated psychotic symptoms (APS)

SIPS Development or Average frequency of at least once Not the effect of substance use Irrespective of current or past
increase by 1 point in per week in the past month and not better explained by a functioning
severity within the past mental disorder
year
CAARMS early Present for at least Frequency of several times per Irrespective of relation to Irrespective of current or past
versions 1 week within the past week substance use or other mental functioning
year and not more than disorders
5 years
CAARMS Symptoms present in At least once a month to twice a Irrespective of relation to 30% drop in SOFAS score from
2006 version the past year week–more than one hour per substance use or other mental premorbid level, sustained for a
occasion disorders month and occurred within
OR past 12 months
At least 3 to 6 times a week–less OR
than one hour per occasion SOFAS score of 50 or less for
past 12 months or longer
Brief limited intermittent, i.e. transient psychotic symptoms (BLIPS)

SIPS Development within Several minutes a day at least 1/ Symptoms are not seriously Irrespective of current or past
the past 3 months month and no more than 1 hour a dangerous or disorganizing, not functioning
day for 4 days a week (on average) the effect of substance use and
for 1 month not better explained by a
mental disorder
CAARMS early Occurrence within the Duration of episode less than a Irrespective of relation to Irrespective of current or past
versions past year week substance use or other mental functioning
disorders
CAARMS Symptoms occurred At least 3 to 6 times a week–more Irrespective of relation to 30% drop in SOFAS score from
2006 version during last year than an hour per occasion substance use or other mental premorbid level, sustained for a
OR disorders month and occurred within
At least daily–less than an hour per past 12 months
occasion OR
SOFAS score of 50 or less for
past 12 months or longer
SOFAS: Social and Occupational Functioning Assessment Scale [3].
sample was part of a larger sample and/or longer follow-up length of follow-up (conversion rates were recorded separately
reported in a study included in the meta-analysis and; at 6-month, 1-year, 2-year, 3-year, 4-year and/or > 4-year
study was only published as an abstract. follow-up where such information was provided; when only
mean sd were provided, the follow-up category next to
2.1.3. Selection process mean + sd was used, e.g., the 3-year follow-up category when
As illustrated in Fig. 2, all titles that turned up in initial searches mean = 26.3 and sd = 9.2 months [62]);
were first examined and assessed for relevance for the main type of CHR criteria (UHR incl. APS, BLIPS and GRFD, COPER and/
question. Next, abstracts of selected papers were examined and or COGDIS) and their distribution;
assessed for relevance and appropriateness of the main question. assessment of UHR criteria (i.e., SIPS, earlier CAARMS versions,
Full texts of potentially relevant papers were obtained and latest CAARMS 2006 version, and other scales (Table 3) such as
independently reviewed by two authors (F.S.L. and C.M.). To rate BSIP, ERIraos or PANSS);
the quality of the studies, we adapted the checklist for assessing sample size and age distribution (age distribution was rated in
the quality of cross-sectional diagnostic accuracy studies by [124] four categories: almost entirely minors ( 18 years; CAD),
to the prospective design of early detection studies. Disagreement almost entirely adults (minimum age 18 years or mean
over inclusion and methodological quality of studies were age > 18 with a lower sd only spanning patients 18 years;
discussed among the two raters until agreement was reached. ADULT), 50% minors (median or mean age 18 years or mean
age 18 with an upper sd still spanning patients 18 years;
2.2. Literature analysis YOUTH), and mixed samples with a proportion of minors
of < 50% (MIX).
2.2.1. Data extraction
For our purpose, we extracted the following variables from the
literature: 2.2.2. Meta-analyses
Analyses and formulae used are specified in the Supplemen-
prevalence of psychosis at follow-up (conversion rates were tary Material S1. In brief, the procedure was as follows: as
recorded separately for CHR criteria where such information was recommended for meta-analyses of univariate studies, propor-
provided). Thereby, the initial (sub)sample size was used as the tions of conversions at follow-up were used as measure of
base rate to avoid a bias towards overly high conversion rates at effect in a fixed-effects model. The inverse variance was used as
longer follow-ups that, even in the absence of additional weight to account for the different sample size of studies.
conversions over time, would result from an increase of drop- Heterogeneity between effect sizes of studies was tested by the
outs/lost-to-follow-ups over time when earlier conversions are Q-statistic, and, in case of significance, a random-effects model
treated as observations-carried-forward; was applied.
Table 3
Overview of instruments used for the assessment of symptomatic ultra-high risk (UHR) criteria.
Instrument Structure and subscales relevant in the rating of APS and BLIPS
SIPS: Structured Interview for Modelled on the ‘Positive and Negative Syndrome Scale’ (PANSS) [41]
Psychosis-Risk Syndromes Contains 4 subscales: positive, negative, disorganized, and general symptoms
[64] Five positive symptoms are used for the assessment of APS and BLIPS: unusual thought content/delusional ideas,
suspiciousness/persecutory ideas, grandiosity, perceptual abnormalities/hallucinations, disorganized communication/speech
CAARMS early (before 2006) Modelled on various scales including the ‘Brief Psychiatric Rating Scale’ (BPRS) [82]
versions: Comprehensive Contains 8 subscales: disorders of thought content, perceptual abnormalities, conceptual disorganization, motor changes,
Assessment of At-Risk Mental concentration and attention, emotion and affect, subjectively impaired energy, and impaired tolerance to normal stress
States [130] Three subscales are used for the assessment of APS and BLIPS: disorder of thought content, perceptual abnormalities,
disorganized speech
CAARMS 2006 version: Modelled on earlier versions of CAARMS [130]

Comprehensive Assessment Contains 7 subscales: positive symptoms, cognitive change attention/concentration, emotional disturbances, negative
of At-Risk Mental States [131] symptoms, behavioural change, motor/physical changes, and general psychopathology
Four positive symptoms are used for the assessment of APS and BLIPS: unusual thought content, non-bizarre ideas, perceptual
abnormalities, and disorganised speech
BSIP: Basel Screening Modelled on the BPRS [82]

Instrument for Psychosis [86] 46-item checklist used in combination with the BPRS
Three symptoms of the BPRS are used for the assessment of APS: hallucinations, unusual thought content, and suspiciousness
Four symptoms of the BPRS are used for the assessment of BLIPS: hallucinations, unusual thought content, suspiciousness, and
conceptual disorganisation
ERIraos: The Early Recognition Modelled on the ‘Instrument for the Retrospective Assessment of the Onset of Schizophrenia’ (IRAOS) [33]
Inventory [63] Consists of a symptom list with 110 items, which is further structured into 12 sections
Five sections include items used for the assessment of APS, BLIPS and also COPER: thought disorder, disorders of self and
delusions, impaired bodily sensations, abnormal perceptions, and observation-based items
PANSS: Positive and Negative Contains 3 subscales: positive, negative and general psychopathology
Syndrome Scale [41] Four positive symptoms are used for the assessment of APS and BLIPS: delusions, hallucinations, suspiciousness, and conceptual
disorganization
BPRS: Brief Psychiatric Rating Contains 24 subscales: somatic concern, anxiety, depression, suicidality, guilt, hostility, elated mood, grandiosity,
Scale [82] suspiciousness, hallucinations, unusual thought content, bizarre behaviour, self-neglect, disorientation, conceptual
disorganisation, blunted affect, emotional withdrawal, motor retardation, tension, uncooperativeness, excitement,
distractibility, motor hyperactivity, mannerisms and posturing
Four subscales of the BPRS are used for the assessment of APS and BLIPS: unusual thought content, hallucinations,
suspiciousness, and conceptual disorganisation
APS: Attenuated Psychotic Symptoms; BLIPS: Brief Limited Intermittent Psychotic Symptoms; COPER: Cognitive-Perceptive basic symptoms.
2.2.3. Sensitivity analyses Guidance Committee and EPA Board (see Acknowledgements) to
To estimate the influence of assessment scales and, relatedly, guarantee that authors had adhered to the consented methodolo-
definitions of UHR criteria (SIPS, CAARMS, CAARMS 2006), type of gy. Only upon its approval by both committees, the manuscript
CHR criteria and combinations (APS, BLIPS, GRFD, COPER, COGDIS, was submitted for external review.
UHR plus COGDIS, UHR and/or COGDIS), and age characteristic of
the sample (CAD, YOUTH or ADULT), the same analyses (Supple- 3. Results
mentary Material S1 and S2) were repeated using these subgroups.
Effect sizes in different subgroups were compared for significant 3.1. Literature search
differences using exploratory one-dimensional x2-tests.
Our preliminary search identified 3467 titles with substantial
2.3. Development of recommendations overlap between the two databases (Fig. 2). After exclusion of titles
published before 1996, the remaining 3054 titles were screened
In line with the EPA’s methodological approach within the and 604 abstracts were examined in more detail for inclusion and
guidance project [28], the consensus process was restricted to the exclusion criteria. Altogether 77 papers were deemed potentially
experts, i.e. authors. General consensus on recommendations was relevant for the meta-analysis and further examined, in particular
achieved by circulating results of the literature search and for likely redundancy of data (i.e., for the inclusion of the sample in
manuscript drafts prepared by the main authors (F.S.-L. and J.K.) a larger sample and/or longer follow-up). From this, 41 papers
to all co-authors for feedback and discussion after the following resulted that were complimented by four additional papers that
steps: reported conversion rates but had a different focus; thus, 45 papers
on 42 samples were finally selected into our meta-analysis (Fig. 2).
compilation of studies to be included in meta-analyses, including
their grade of evidence rating; 3.2. Included studies and study design
conducting of analyses and first drafting of the manuscript and;
recurrently adaptating after each feedback-related until full Supplementary Table 1 gives the description of included studies
agreement among authors was reached on the manuscript’s that generally were rated on levels of evidence according to the
submission version. This step was also performed once more Scottish Intercollegiate Guideline Network (SIGN) as ‘2+’ (i.e.,
after receiving external review. cohort studies with a low risk of confounding, bias, or chance and a
moderate probability that the relationship is causal). Seven
Furthermore, during the process of guidance development, the samples (16.7%) each were from North America [1,12,14,15,42,
manuscript’s submission version underwent review by the EPA 100,126] and Australia [8,62,77–79,123,132,133], six (14.3%) from
Table 4
Basic symptom criteria.
Cognitive-Perceptive Basic Symptoms (COPER)
Presence of 1 of the following 10 basic symptoms with a SPI-A score
of 3 within the last 3 months and first occurrence 12 months ago
Thought interferencea
Thought perseveration
Thought pressurea
Thought blockagesa
Disturbance of receptive speecha
Decreased ability to discriminate between ideas/perception, fantasy/true
memories
Unstable ideas of referencea
Derealisation
Visual perception disturbances (excl. hypersensitivity to light and blurred
vision)
Acoustic perception disturbances (excl. hypersensitivity to sounds)
Cognitive Disturbances (COGDIS)
Presence of 2 of the following 9 basic symptoms with a SPI-A score
of 3 within the last 3 months
Inability to divide attention
Thought interferencea
Thought pressurea Fig. 2. Flow of studies retrieved by the systematic literature search with the
Thought blockagesa algorithm detailed in 2.1.1. * published before 1996, i.e. before the first description
Disturbance of receptive speecha of current UHR criteria (Yung and McGorry, 1996) [128].
Disturbance of expressive speech
Unstable ideas of referencea
Disturbances of abstract thinking
Captivation of attention by details of the visual field conversion rates for the three UHR criteria were reported. For the
SPI-A: Schizophrenia Proneness Instrument, Adult version [104].
identification of a CHR according to basic symptoms criteria, the
a
Indicates basic symptoms included in both COPER and COGDIS. ‘Bonn Scale for the Assessment of Basic Symptoms’ (BSABS; [32])
was used in two (4.8%) samples (in one in combination with the
CAARMS), the SPI-A in six (14.3%; in five in combination with the
Germany [9,11,50,55,91,103,105,112], three (7.1%) each from the SIPS) and the ERIraos in one (2.4%). Assessments were generally
Netherlands [116,120,135] and the UK [26,69–72], two (4.8%) each carried out by clinicians specifically trained in the application of
from Switzerland [87,113] and Finland [59,60], and six (14.3%) the respective instruments.
each from other European [5,24,48,53,58,92] and from Asian The majority of information on conversion to psychosis was
countries [40,45,52,56,57,134]. All but one [60] were help-seeking available for the 2-year follow-up (on 23 samples; 54.8%, incl.
clinical samples: 25 (59.5%) of established early detection and reports on 18-month conversion rates). Conversion rates at
intervention (EDI) services and 16 (38.1%) recruited in mental 6 months were reported for ten (23.8%) samples and at 12 months
health services for the purpose of an EDI study. CHR patients for 20 (47.6%). Information on 3-year conversion rates (incl. reports
generally received some kind of treatment, yet for the six (14.3%) on 2.5-year conversion rates) was obtained from ten (23.8%), on 4-
treatment trials included in the meta-analysis, we considered only year conversion rates from eight (19.0%) and on longer follow-ups
the control groups [1,5,9,69–72,116]. Six (14.3%) CHR samples from five (11.9%) samples.
were complimented by a CHR-negative group [40,59,60,112, Conversion was mainly to a non-affective schizophrenic or
113,132,133]. schizophreniform psychosis according to DSM-IV (71.4–100% of
Baseline CHR sample sizes ranged from seven [60] to 817 [77] converters). A conversion to an affective psychosis was generally
including altogether 4952 CHR subjects. With regard to age at rare, with reported rates in converters between 2 and 28.6%. For 15
baseline, the age range of CHR samples was almost always (35.7%) samples, information on type of psychosis was not
somewhere between 12 and 40 years; one sample [56] included provided.
patients as young as 6 years, and one other [50] patients as old as
53 years. Fifteen (35.7%) samples almost entirely included adults, 3.3. Heterogeneity analyses of fixed-effects models
ten (23.8%) a majority of patients aged 18 years, six (14.3%)
almost entirely minors, and 10 (23.8%) mixed samples with a The Q-statistic indicated that there was significant heteroge-
dominance of adults; one study lacked information on age [48]. In neity between the conversion rate estimates in UHR, COPER and
28 samples (66.7%), 50% were males; two (4.8%) studies did not COGDIS samples at the different follow-ups with the exception of
provide data on gender distribution [48,53]. On 22 (52.4%) samples the 1-year conversion rates of two COGDIS studies (Q(1) = 0.10;
information on co-morbidities at baseline was provided that were I2 = 39%). In all other cases, I2 was between 54% and 98% and,
mainly affective disorders (23.1–75.8%) and anxiety disorders consequently, E s and their 95% CIs were calculated according to
(8.7–57.6%). the random-effects model.
Participation rates of eligible samples at baseline were reported
for 24 (57.1%) samples (39.1–100%); the drop-out or non- 3.4. Conversion rates to psychosis
participation rates at last follow-up reported for 36 (85.7%)
samples were between 0 and 91.6%. Overall, the pooled conversion rate in UHR samples increased
For the identification of an CHR according to UHR criteria, in 17 from 9.6% at 6 months to 37.0% at > 4-year follow-up (Fig. 3a–e). In
(40.5%) samples the SIPS was employed, in twelve (28.6%) an early COGDIS samples, the respective numbers ranged from 25.3% at
CAARMS version (of before 2006; in two studies [77,79], CAARMS 1 year to 61.3% at > 4 years; a 6-month conversion rate of 13.9%
and BRPS data were mixed), in six (14.3%) the latest 2006 CAARMS was only reported by one study, thus not allowing the calculation
version including the general obligate functional decline criterion of E (Fig. 4). For COPER, E and E could be calculated for 1- and
(Tab. 2), in two (4.8%) each the PANSS and ERIraos, and in one 2-year follow-ups for that they were 14.4% and 21.1% (Supple-
(2.4%) the BSIP. For eleven (26.2%) UHR samples differential mentary Table 2). Overall, z-values of E=E were > 2.58 indicating
Fig. 3. a–e Conversion rates at different follow-ups in samples meeting any one ‘Ultra-high risk’ (UHR) criterion (irrespective of the potential presence of basic symptom
criteria). Upper number indicates scale used for the assessment of UHR criteria (1: SIPS; 2: CAARMS; 3: CAARMS 2006 version; and 4: other scale). Upper small letter indicates
age group of sample (a: ADULT; b: MIX; c: YOUTH; and d: CAD). F: according to fixed-effects model.
in APS samples was not available. Data on conversion rates in BLIPS

samples were even less available (Supplementary Table 2), and
pooled conversion rates could only be calculated for 2-year and 3-
year follow-ups with a significant pooled sample effect only at
3 years due to the 3 BLIPS-non-converters [113] at 2 years. Pooled
2- and 3-year conversion rates in GRFD samples (Supplementary
Table 2) were much lower, showed no significant pooled sample
effect, and were equal to or even lower than conversion rates of
CHR-negative samples (Supplementary Table 3).
Pooled conversion rates between the three UHR criteria differed
significantly at 1-, 2- and 3-year follow-ups (x2ð2Þ 15:200;
P < 0.001), mainly due to higher conversion rates in BLIPS and/
or lower in GRFD samples (Supplementary Table 3).
Since recent studies have suggested that the combined
assessment of UHR and BS criteria, especially COGDIS, was
advantageous to their exclusive assessment in identifying an
CHR [92,112], we also analyzed pooled effects of both ‘UHR plus
COGDIS’ and ‘UHR and/or COGDIS’ on 2-year conversion rates. For
both combinations, most conversion rates of studies were
significantly heterogeneous; and pooled conversion rates of
random-effects models were 26.7% for ‘UHR plus COGDIS’ and
19.9% for ‘UHR and/or COGDIS’ (Supplementary Table 2); both
indicated a significant pooled sample effect that did not
significantly differ from each other (x2ð1Þ ¼ 0:992; P > 0.25) or
from UHR (x2ð2Þ ¼ 1:461; P > 0.25) or COGDIS (x2ð2Þ ¼ 1:618;
P > 0.25) when these were considered irrespective but not
exclusive of each other as in the two earlier studies [92,112].
Fig. 4. Conversion rates at different follow-ups in samples meeting the ‘Cognitive

disturbances’ (COGDIS) criterion (irrespective of the potential presence of COPER or
3.5.2. UHR criteria assessment scales
UHR criteria). Upper number indicates scale used for the assessment of COGDIS (5: For the differences in requirements of onset, recency,
BSABS; and 6: SPI-A). Upper small letter indicates age group of sample (a: ADULT; frequency and psychosocial functioning in symptomatic UHR
and d: CAD). F: according to fixed-effects model. criteria of different scales [110] (Table 2), conversion rates in UHR
samples were additionally calculated separately for SIPS-, early
CAARMS- and CAARMS 2006-assessed samples (Fig. 3a–e;
significant pooled sample effect of UHR, COPER and COGDIS Supplementary Table 4). Overall, there was no indication of a
samples on conversion rates to psychosis at an at least 5% error significant effect of the applied scale and, relatedly, UHR
level. definition; only at 4-year follow-up, there was some weak
A significant pooled sample effect was mainly missing in the indication of a difference between UHR assessments
samples not considered at CHR (Supplementary Table 2) that (x2ð2Þ ¼ 6:416; P < 0.05) due to a lower single effect of CAARMS
showed mainly homogeneous conversion rates; of these, only the 2006 [53] in comparison with the single effect of CAARMS early
9.8% conversion rate at > 4-year follow-up revealed a significant versions [79] (Supplementary Table 4).
pooled sample effect. Thus, compared to even the lowest
conversion rate in UHR, COPER or COGDIS samples at any 3.5.3. Age group
follow-up (Figs. 3 and 4; Supplementary Table 2), conversion To examine a potential age effect in UHR samples, we compared
rates of help-seeking patients not meeting the examined CHR three age groups (CAD, YOUTH and ADULT) for available
criteria were clearly significantly lower (x2ð1Þ 5:175; P < 0.025). conversion rates on 6-month, 1-, 2- and > 4-year follow-ups
Irrespective of a potential co-occurrence of criteria, UHR, COPER (Fig. 3a–c,e). Pairwise comparisons indicated lower conversion
and COGDIS samples did not significantly differ in 6-month, 1- and rates in CAD compared to YOUTH throughout, and additionally to
2-year conversion rates (x2ð2Þ 4:118; P > 0.10) but only in longer ADULT at 2 and > 4 years (Supplementary Table 5). YOUTH
term conversion rates (x2ð2Þ 6:767; P < 0.05) due to higher rates in conversion rates never differed significantly from those in ADULT.
both COPER and COGDIS compared to UHR samples (x2ð1Þ 5:522; The only significant difference to the total sample conversion rates
P < 0.05). occurred for CAD at > 4 years with an additional trend result at
6 months and 2 years (Supplementary Table 5).
3.5. Sensitivity analyses
3.5.1. Single and combined CHR criteria 4. Recommendations

With regard to the analyses of UHR studies reporting
conversion rates separately for the three UHR criteria (APS, BLIPS 4.1. Meta-analysis of studies as the evidence base of the European
and GRFD), except for the conversion rates of APS at 3 and that of Guidance
BLIPS at 2 years (Supplementary Table 2), fixed-effects models
were chosen for their non-significant Q-statistic or negative t2- Based on the results of our meta-analyses, we found that
value. recommendations can be formulated with sufficient evidence
As detailed in Supplementary Table 2, the pooled conversion based on studies mainly given SIGN ‘2+’ rating (due to the
rate in APS samples ranged from 7.7% at 6 months to 14.9% unfeasibility of RCTs in early detection research) at a grade of
at > 4-year follow-ups of CAD samples [15,60]; and all pooled recommendation of ‘C’ for recommendations 1–5 and at grade ‘D’
sample effects were significant. Data on the 4-year conversion rate for the expert consensus-based recommendation 6.
4.2. Proposed recommendations of the European Guidance Project 4.2.4. Recommendation 4

The EPA considers that the above CHR criteria should only be
4.2.1. Recommendation 1 applied in persons already distressed by mental problems and
The EPA considers that the following three CHR criteria should seeking help for them or persons seeking clarification of their
be alternatively used in the early detection of psychosis when past current risk for a vulnerability for psychosis, e.g., by genetic risk.
or present psychosis and causation by a somatic illness had been Any clinical screening of other persons seems not warranted by
ruled out: current scientific evidence.
at least any one attenuated psychotic symptom (i.e., (1) unusual 4.2.5. Recommendation 5
thought contents or delusional ideas not held with full The EPA considers that the above CHR criteria should only be
conviction, including ideas of reference not immediately rectified used and communicated with outmost care in children and young
by cognition, (2) perceptual aberrations or hallucination with adolescents in whom they should nevertheless be assessed and
remaining insight, or (3) disorganized communication or speech monitored (see Schmidt SJ et al.; this issue). In late adolescence,
that is still comprehensible and responds to structuring in the however, the CHR criteria seem to be as applicable as in adults.
interview) that meets the additional requirements of either SIPS
or early CAARMS (Table 2); 4.2.6. Recommendation 6
at least any two self-experienced and self-reported cognitive The EPA considers that a trained specialist (psychiatrist, clinical
basic symptoms rated irrespective of their appearance in the psychologist or equivalent mental health professional) with
interview (i.e., (1) interference of completely insignificant sufficient experience in CHR should carry out the assessment; if
thought contents, (2) blockage of thoughts not explained by referral to a specialist is not possible, the responsible clinician
lack of concentration or attention, (3) thought pressure by should consult a trained specialist on the case, e.g. by phone; and
thoughts unrelated to a common topic, (4,5) disturbances of specialized early detection services should be prepared to give
receptive or expressive speech in everyday use of native such advice, e.g., within the framework of telephone consultation
language, (6) inability to divide attention between tasks relating hours. Case conferences with experts in early detection of
to different senses and generally not requiring full attention each psychoses are even advised for mental health specialists.
such as making a sandwich and talking to someone, (7)
disturbance in the immediate recognition and understanding
of any kind of abstract, figurative or symbolic phrases or 5. Discussion
contents, (8) subjective experience of self-reference that are
almost immediately rectified by cognition, and (9) captivation of From our meta-analyses, evidence-based recommendations for
attention by insignificant details of the visual field that impairs early detection of psychosis were formulated that improve upon
paying attention to more relevant stimuli) that have not been those of previous expert consensus guidelines [10,20,74–76]. While
present in what the patient considers his/her premorbid stage, the evidence for the psychosis-predictive value of UHR criteria,
have occurred at least on a weekly basis for some time in the past especially APS and BLIPS, and basic symptom criteria, especially
3 months and are not an effect of drug use; COGDIS, continue to accumulate, the heterogeneity of conversion
at least any one transient psychotic symptom (i.e., delusion, rates between CHR samples strongly suggests the presence of
hallucination, formal thought disorder) that meets the additional moderating variables. Of these, single CHR criteria, their assessment
requirements of either SIPS or early CAARMS (Table 2). mode and definition, and age were analyzed based on the limited
available data. Generally, the lack of detailed information on the
relationship of sample, presence and kind of treatment and study
4.2.2. Recommendation 2 characteristics, and clinical variables with conversion rates impeded
The EPA considers that a genetically increased risk of psychosis our analyses. Thus, for example, the frequent specification of simply
by a positive family history of psychosis in at least one first-degree mean follow-up times along with the frequent lack of time-
biological relative should not be used as a clinical indicator of a dependent survival analyses only allowed estimates of the impact of
CHR on its own, even if accompanied by functional deficits and observation time on conversion rate. This lack of information also
mental problems. Rather, it should be regarded as a general risk precluded the simultaneous consideration of our considered
factor indicating an already increased pre-CHR assessment risk for moderators and might have introduced some bias. For example,
psychosis that should be taken into account in patients meeting the pooled conversion rate in APS samples with > 4-year follow-up
the above CHR criteria. Patients not presenting the above CHR was lower than at 2- or 3-year follow-up, yet these shorter follow-
criteria but a genetic risk and other mental problems should ups were in older samples than the > 4-year follow-ups on that data
however be encouraged to present again for a CHR assessment were provided only by CAD samples, and significantly lower
should they note the onset of mental problems resembling CHR conversion rates can be assumed for this age range. Furthermore,
symptoms. our estimates might somewhat underestimate the true effect of a
CHR on conversion to psychosis as we conservatively treated the
4.2.3. Recommendation 3 often considerable numbers of drop-outs and/or cases with shorter
In line with the general EPA guidance on prevention of mental than the allocated follow-up as non-converters. Despite these
disorders [13] whose aims include reduction of the burden of shortcomings, some consistent patterns occurred that allowed the
mental disorders by improvement in quality of life and productiv- formulation of detailed recommendations including age consider-
ity of individuals, the EPA considers that a significant decline in ations for the first time.
occupational and/or social functioning (and, relatedly, in produc- These recommendations did not include the basic symptom
tivity) should not be an obligate requirement in the above CHR criterion COPER because of its large overlap with COGDIS and,
criteria for the lack of evidence for an improvement of prediction compared to COGDIS, its lesser degree of evidence due to the very
by this addition. However, it should be considered as an indication small number of studies of that only the study on that the basic
of an imminence of risk of conversion and CHR patients with a symptom criteria were developed [50,103] exceeded 2-year
significant functional decline should be considered at high need for follow-up. Moreover, the recommendations did not adopt the
treatment. recent addition of an obligate functional decline criterion to the
symptomatic UHR criteria, as our analyses did not support the References
presumption that it would improve prediction of psychosis in help-
seeking samples. The addition had been put forward on the basis of [1] Addington J, Epstein I, Liu L, French P, Boydell KM, Zipursky RB. A randomized
controlled trial of cognitive behavioral therapy for individuals at clinical high
consistent reports of significantly lower functioning in converters in
risk of psychosis. Schizophr Res 2011;125(1):54–61.
group mean comparisons and the frequent inclusion of functional [2] Alaghband-Rad J, McKenna K, Gordon CT, Albus KE, Hamburger SD, Rumsey
deficits in prediction models [14,17,21,30,39,62,92,93,119,121, JM, et al. Childhood-onset schizophrenia: the severity of premorbid course. J
129,136]. However, such group mean-based results seem to fail Am Acad Child Adolesc Psychiatry 1995;34(10):1273–83.
[3] American Psychiatric Association. Diagnostic and statistical manual of mental
to translate into an improved prediction in practice, highlighting the disorders (DSM-IV), 4th ed., Washington, DC: American Psychiatric Press;
general need for more translational research on predictors using, for 1994.
example, existing norms or clearly defined and tested cut-offs [4] American Psychiatric Association. Diagnostic and statistical manual of mental
disorders (DSM-5), 5th ed., Washington, DC: American Psychiatric Press;
[67,101]. Furthermore, a ‘one-fits-all’ approach most likely does not 2013.
measure up to the considerable heterogeneity of conversion rates [5] Amminger GP, Schäfer MR, Papageorgiou K, Klier CM, Cotton SM, Harrigan
even in CHR samples of equal intake criteria. Future early detection SM, et al. Long-chain omega-3 fatty acids for indicated prevention of psy-
chotic disorders: a randomized, placebo-controlled trial. Arch Gen Psychiatry
approaches should therefore define different CHR groups that are 2010;67(2):146–54.
identified, for example, by a risk stratification approach, which [6] Ballageer T, Malla A, Manchanda R, Takhar J, Haricharan R. Is adolescent-onset
might consider most likely level of functioning but also other first-episode psychosis different from adult onset? J Am Acad Child Adolesc
Psychiatry 2005;44(8):782–9.
potential predictors such as neurocognitive or neurobiological
[7] Bartels-Velthuis AA, van de Willige G, Jenner JA, van Os J, Wiersma D. Course
abnormalities [5,11,54,67,80,87]. of auditory vocal hallucinations in childhood: 5-year follow-up study. Br J
Psychiatry 2011;199(4):296–302.
6. Conclusions and perspectives [8] Bechdolf A, Thompson A, Nelson B, Cotton S, Simmons MB, Amminger GP,
et al. Experience of trauma and conversion to psychosis in an ultra-high-risk
(prodromal) group. Acta Psychiatr Scand 2010;121(5):377–84.
The young field of preventive research in psychosis has already [9] Bechdolf A, Wagner M, Ruhrmann S, Harrigan S, Putzfeld V, Pukrop R, et al.
resulted in sufficient evidence to formulate recommendations for Preventing progression to first-episode psychosis in early initial prodromal
states. Br J Psychiatry 2012;200(1):22–9.
an early detection of psychosis in the clinical practice. Yet, our [10] Bertolote J, McGorry P. Early intervention and recovery for young people with
analysis has also revealed significant heterogeneity of conversion early psychosis: consensus statement. Br J Psychiatry 2005;48(Suppl.):116–
rates that needs to be addressed in future studies in order to 9.
[11] Bodatsch M, Ruhrmann S, Wagner M, Müller R, Schultze-Lutter F, Frommann
develop more sophisticated prediction models that can be easily I, et al. Prediction of psychosis by mismatch negativity. Biol Psychiatry
translated into clinical practice and address the special character- 2011;69(10):959–66.
istics and treatment needs of different patient groups. Further- [12] Buchy L, Perkins D, Woods SW, Liu L, Addington J. Impact of substance use on
conversion to psychosis in youth at clinical high risk of psychosis. Schizophr
more, the success of preventive approaches also depends on a
Res 2014;156(2–3):277–80.
sufficiently high rate of target persons who are reached by it. Thus, [13] Campion J, Bhui K, Bhugra D, European Psychiatric Association. European
to reach also CHR persons who do not actively seek help for their Psychiatric Association (EPA) guidance on prevention of mental disorders.
Eur Psychiatry 2012;27(2):68–80.
mental problems, more research in the general population is
[14] Cannon TD, Cadenhead K, Cornblatt B, Woods SW, Addington J, Walker E, et al.
needed to develop ethically justified means such as well-validated Prediction of psychosis in youth at high clinical risk: a multisite longitudinal
and reliable screeners [68] on the basis of those already proposed study in North America. Arch Gen Psychiatry 2008;65(1):28–37.
[49,66,85]. [15] Carrión RE, McLaughlin D, Goldberg TE, Auther AM, Olsen RH, Olvet DM, et al.
Prediction of functional outcome in individuals at clinical high risk for
psychosis. JAMA Psychiatry 2013;70(11):1133–42.
Disclosure of interest [16] Collins PY, Patel V, Joestl SS, March D, Insel TR, Daar AS, et al. Grand challenges
in global mental health. Nature 2011;475(7354):27–30.
[17] Cornblatt BA, Carrión RE, Addington J, Seidman L, Walker EF, Cannon TD, et al.
The authors declare that they have no conflicts of interest Risk factors for psychosis: impaired social and role functioning. Schizophr
concerning this article. Bull 2012;38(6):1247–57.
[18] David CN, Greenstein D, Clasen L, Gochman P, Miller R, Tossell JW, et al.
Childhood onset schizophrenia: high rate of visual hallucinations. J Am Acad
Acknowledgements Child Adolesc Psychiatry 2011;50(7):681–6.
[19] De Girolamo G, Dagani J, Purcell R, Cocchi A, McGorry PD. Age of onset of
mental disorders and use of mental health services: needs, opportunities and
Obtaining of required approval of EPA Guidance Committee and obstacles. Epidemiol Psychiatr Sci 2012;21(1):47–57.
EPA Board was coordinated by the EPA President, Wolfgang Gaebel [20] Deutsche Gesellschaft für Psychiatrie, Psychotherapie und Nervenheilkunde
(Germany). In alphabetical order, EPA Guidance Committee (DGPPN). Behandlungsleitlinie Schizophrenie. Darmstadt: Steinkopff Verlag;
2006.
members were: Dinesh Bhugra (UK), Peter Falkai (Germany), [21] Dragt S, Nieman DH, Veltman D, Becker HE, van de Fliert R, de Haan L, et al.
Andrea Fiorillo (Italy), Wolfgang Gaebel, Reinhard Heun (UK), Environmental factors and social adjustment as predictors of a first psychosis
Hans-Jürgen Möller (Germany), Michael Musalek (Austria) and in subjects at ultra high risk. Schizophr Res 2011;125(1):69–76.
[22] The European Network of Schizophrenia Networks for the Study of Gene-
Danuta Wasserman (Sweden); EPA Board members were: Sue Environment Interactions (EU-GEI). Schizophrenia aetiology: do gene-
Bailey (UK), Julian Beezhold (UK), Geert Dom (Belgium), Peter environment interactions hold the key? Schizophr Res 2008;102(1–3):
Falkai, Andrea Fiorillo, Wolfgang Gaebel, Silvana Galderisi (Italy), 21–6.
[23] Fiori Nastro P, Schimmelmann BG, Gebhardt E, Monducci E, Resch F, Koch E,
Paz Garcı́a-Portilla (Spain), Philip Gorwood (France), Cécile Hanon
et al. [Challenges in the early detection of psychosis in children and adoles-
(France), Andreas Heinz (Germany), Marianne Kastrup (Denmark), cents]. Riv Psichiatr 2012;47(2):116–25.
Levent Küey (Turkey), Tamas Kurimay (Hungary), Michael [24] Fusar-Poli P, Hobson R, Raduelli M, Balottin U. Reliability and validity of the
Musalek, Wulf Rössler (Switzland), Jerzy Samochowiec (Poland), comprehensive assessment of the at risk mental state, Italian version
(CAARMS-I). Curr Pharm Des 2012;18(4):386–91.
Rutger J. van der Gaag (Netherlands) and Danuta Wasserman [25] Fusar-Poli P, Borgwardt S, Bechdolf A, Addington J, Riecher-Rössler A,
(source: www.europsy.net). Schultze-Lutter F, et al. The psychosis high-risk state: a comprehensive
state-of-the-art review. JAMA Psychiatry 2013;70(1):107–20.
[26] Fusar-Poli P, Byrne M, Badger S, Valmaggia LR, McGuire PK. Outreach and
Appendix A. Supplementary data support in south London (OASIS), 2001–2011: ten years of early diagnosis and
treatment for young individuals at high clinical risk for psychosis. Eur
Psychiatry 2013;28(5):315–26.
Supplementary data associated with this article can be found, in [27] Fux L, Walger P, Schimmelmann BG, Schultze-Lutter F. The schizophrenia
the online version, at http://dx.doi.org/10.1016/j.eurpsy.2015.01. proneness instrument, child and youth version (SPI-CY): practicability and
010. discriminative validity. Schizophr Res 2013;146(1–3):69–78.
[28] Gaebel W, Möller HJ. European guidance–a project of the European Psychi- [55] Koutsouleris N, Meisenzahl EM, Davatzikos C, Bottlender R, Frodl T, Scheuer-
atric Association. Eur Psychiatry 2012;27:65–7. ecker J, et al. Use of neuroanatomical pattern classification to identify
[29] Gater R, Jordanova V, Maric N, Alikaj V, Bajs M, Cavic T, et al. Pathways to subjects in at-risk mental states of psychosis and predict disease transition.
psychiatric care in Eastern Europe. Br J Psychiatry 2005;186:529–35. Arch Gen Psychiatry 2009;66(7):700–12.
[30] González-Pinto A, Ruiz de Azúa S, Ibáñez B, Otero-Cuesta S, Castro-Fornieles J, [56] Lam MM, Hung SF, Chen EY. Transition to psychosis: 6-month follow-up of a
Graell-Berna M, et al. Can positive family factors be protective against the Chinese high-risk group in Hong Kong. Aust N Z J Psychiatry 2006;40(5):414–20.
development of psychosis? Psychiatry Res 2011;186(1):28–33. [57] Lee J, Rekhi G, Mitter N, Bong YL, Kraus MS, Lam M, et al. The longitudinal
[31] Gore FM, Bloem PJ, Patton GC, Ferguson J, Joseph V, Coffey C, et al. Global youth at risk study (LYRIKS)–an Asian UHR perspective. Schizophr Res
burden of disease in young people aged 10–24 years: a systematic analysis. 2013;151(1–3):279–83.
Lancet 2011;377(9783):2093–102. [58] Lemos-Giráldez S, Vallina-Fernández O, Fernández-Iglesias P, Vallejo-Seco G,
[32] Gross G, Huber G, Klosterkötter J, Linz M. Bonner Skala für die Beurteilung von Fonseca-Pedrero E, Paı́no-Piñeiro M, et al. Symptomatic and functional
Basisymptomen.BSABS; Bonn Scale for the assessment of basic symptoms outcome in youth at ultra-high risk for psychosis: a longitudinal study.
Berlin: Springer-Verlag; 1987. Schizophr Res 2009;115(2–3):121–9.
[33] Häfner H, Riecher-Rössler A, Hambrecht M, Maurer K, Meissner S, Schmidtke [59] Lindgren M, Manninen M, Kalska H, Mustonen U, Laajasalo T, Moilanen K, et al.
A, et al. IRAOS: an instrument for the assessment of onset and early course of Predicting psychosis in a general adolescent psychiatric sample. Schizophr Res
schizophrenia. Schizophr Res 1992;6(3):209–23. 2014;158(1–3):1–6. http://dx.doi.org/10.1016/j.schres.2014.06.028.
[34] Häfner H, Maurer K, Löffler W, Riecher-Rössler A. The influence of age and sex on [60] Manninen M, Lindgren M, Therman S, Huttunen M, Ebeling H, Moilanen I,
the onset and early course of schizophrenia. Br J Psychiatry 1993;162:80–6. et al. Clinical high-risk state does not predict later psychosis in a delinquent
[35] Häfner H, Maurer K, Löffler W, Fätkenheuer B, an der Heiden W, Riecher- adolescent population. Early Interv Psychiatry 2014;8(1):87–90.
Rössler A, et al. The epidemiology of early schizophrenia. Influence of age and [61] Marshall M, Lewis S, Lockwood A, Drake R, Jones P, Croudace T. Association
gender on onset and early course. Br J Psychiatry 1994;Suppl.(23):29–38. between duration of untreated psychosis and outcome in cohorts of first-
[36] Harrison PJ, Weinberger DR. Schizophrenia genes, gene expression and episode patients: a systematic review. Arch Gen Psychiatry 2005;62 (9):975–
neuropathology on the matter of their convergence. Mol Psychiatry 83.
2005;10(Suppl. 3):40–68. [62] Mason O, Startup M, Halpin S, Schall U, Conrad A, Carr V. Risk factors for
[37] Hlastala SA, McClellan J. Phenomenology and diagnostic stability of youths transition to first episode psychosis among individuals with ‘at-risk mental
with atypical psychotic symptoms. J Child Adolesc Psychopharmacol 2005;15 states’. Schizophr Res 2004;71(2–3):227–37.
(3):497–509. [63] Maurer K, Hörrmann F, Trendler G, Schmidt M, Haefner H. [Identification of
[38] Jackson D, Kirkbride J, Croudace T, Morgan C, Boydell J, Errazuriz A, et al. psychosis risk by the Early Recognition Inventory (ERIraos) – Description of
Meta-analytic approaches to determine gender differences in the age-inci- the schedules and preliminary results on reliability and validity of the
dence characteristics of schizophrenia and related psychoses. Int J Methods checklist]. Nervenheilkunde 2006;25(1):11–6.
Psychiatr Res 2013;22(1):36–45. [64] McGlashan T, Walsh B, Woods S. The psychosis-risk syndrome. Handbook for
[39] Jang JH, Shin NY, Shim G, Park HY, Kim E, Jang GE, et al. Longitudinal patterns diagnosis and follow-up. New York, NY: Oxford University Press; 2010.
of social functioning and conversion to psychosis in subjects at ultra-high [65] McGrath J, Saha S, Welham J, El SO, MacCauley C, Chant D. A systematic
risk. Aust N Z J Psychiatry 2011;45(9):763–70. review of the incidence of schizophrenia: the distribution of rates and the
[40] Katsura M, Ohmuro N, Obara C, Kikuchi T, Ito F, Miyakoshi T, et al. A influence of sex, urbanicity, migrant status and methodology. BMC Med
naturalistic longitudinal study of at-risk mental state with a 2.4-year fol- 2004;2:13.
low-up at a specialized clinic setting in Japan. Schizophr Res 2014;158(1– [66] Meneghelli A, Alpi A, Cascio MT, Häfner H, Maurer K, Preti A, et al. Italian
3):32–8. http://dx.doi.org/10.1016/j.schres.2014.06.013. validation of the early recognition inventory for the retrospective assessment
[41] Kay SR, Fiszbein A, Opfer LA. The positive and negative syndrome scale of the onset of schizophrenia checklist: reliability, validity and instructions
(PANSS) for schizophrenia. Schizophr Bull 1987;13(2):261–76. for use. J Psychopathol 2013;19(1–2):1–9.
[42] Kayser J, Tenke CE, Kroppmann CJ, Alschuler DM, Ben-David S, Fekri S, et al. [67] Michel C, Ruhrmann S, Schimmelmann BG, Klosterkötter J, Schultze-Lutter F.
Olfaction in the psychosis prodrome: electrophysiological and behavioral A stratified model for psychosis prediction in clinical practice. Schizophr Bull
measures of odor detection. Int J Psychophysiol 2013;90(2):190–206. 40(6):1533–42, doi:10.1093/schbul/sbu025.
[43] Kelleher I, Murtagh A, Molloy C, Roddy S, Clarke MC, Harley M, et al. [68] Michel C, Schultze-Lutter F, Schimmelmann BG. Screening instruments in
Identification and characterization of prodromal risk syndromes in young child and adolescent psychiatry: general and methodological considerations.
adolescents in the community: a population-based clinical interview study. Eur Child Adolesc Psychiatry 23(9):725–7, doi:10.1007/s00787-014-0608-x.
Schizophr Bull 2012;38(2):239–46. [69] Morrison AP, French P, Walford L, Lewis SW, Kilcommons A, Green J, et al.
[44] Kelleher I, Keeley H, Corcoran P, Lynch F, Fitzpatrick C, Devlin N, et al. Cognitive therapy for the prevention of psychosis in people at ultra-high risk:
Clinicopathological significance of psychotic experiences in non-psychotic randomised controlled trial. Br J Psychiatry 2004;185:291–7.
young people: evidence from four population-based studies. Br J Psychiatry [70] Morrison AP, French P, Parker S, Roberts M, Stevens H, Bentall RP, et al. Three-
2012;201(1):26–32. year follow-up of a randomized controlled trial of cognitive therapy for the
[45] Kim E, Jang JH, Park HY, Shim G, Hwang JY, Kim SN, et al. Pharmacotherapy prevention of psychosis in people at ultrahigh risk. Schizophr Bull
and clinical characteristics of ultra-high-risk for psychosis according to 2007;33(3):682–7.
conversion status: a naturalistic observational study. Early Interv Psychiatry [71] Morrison AP, Stewart SL, French P, Bentall RP, Birchwood M, Byrne R, et al.
2012;6(1):30–7. Early detection and intervention evaluation for people at high-risk of psy-
[46] Kirkbride JB, Fearon P, Morgan C, Dazzan P, Morgan K, Tarrant J, et al. chosis-2 (EDIE-2): trial rationale, design and baseline characteristics. Early
Heterogeneity in incidence rates of schizophrenia and other psychotic syn- Interv Psychiatry 2011;5(1):24–32.
dromes: findings from the 3-center AeSOP study. Arch Gen Psychiatry [72] Morrison AP, French P, Stewart SL, Birchwood M, Fowler D, Gumley AI, et al.
2006;63(3):250–8. Early detection and intervention evaluation for people at risk of psychosis:
[47] Kirkbride JB, Errazuriz A, Croudace TJ, Morgan C, Jackson D, Boydell J, et al. multisite randomised controlled trial. BMJ 2012;344:e2233.
Incidence of schizophrenia and other psychoses in England, 1950–2009: a [73] Mrazek PJ, Haggerty HJ. Reducing risks for mental disorders: frontiers for
systematic review and meta-analyses. PLoS One 2012;7(3):e31660. preventive research. Washington, DC: Academy Press; 2013.
[48] Kiss I, Kelemen O, Kéri S. Decreased peripheral expression of neuregulin 1 in [74] National Institute for Health, Clinical Excellence (NICE). Psychosis with
high-risk individuals who later converted to psychosis. Schizophr Res coexisting substance misuse: assessment and management in adults and
2012;135(1–3):198–9. young people;[Online] 2011 [Available at: https://www.nice.org.uk/
[49] Kline E, Schiffman J. Psychosis risk screening: a systematic review. Schizophr guidance/published?type=Guidelines (Accessed: 25 August 2014)].
Res 2014;158(1–3):11–8. http://dx.doi.org/10.1016/j.schres.2014.06.036. [75] National Institute for Health, Clinical Excellence (NICE). Psychosis and schizo-
[50] Klosterkötter J, Hellmich M, Steinmeyer EM, Schultze-Lutter F. Diagnosing phrenia in children and young people: recognition and management;[Online]
schizophrenia in the initial prodromal phase. Arch Gen Psychiatry 2001;58 2013 [Available at: https://www.nice.org.uk/guidance/published?type=
(2):158–64. Guidelines (Accessed: 25 August 2014)].
[51] Klosterkötter J, Schultze-Lutter F, Bechdolf A, Ruhrmann S. Prediction and [76] National Institute for Health, Clinical Excellence (NICE). Psychosis and schizo-
prevention of schizophrenia: what has been achieved and where to go next? phrenia in adults: treatment and management;[Online] 2014 [Available at:
World Psychiatry 2011;10(3):165–74. https://www.nice.org.uk/guidance/published?type=Guidelines (Accessed:
[52] Koike S, Takano Y, Iwashiro N, Satomura Y, Suga M, Nagai T, et al. A multimodal 25 August 2014)].
approach to investigate biomarkers for psychosis in a clinical setting: the [77] Nelson B, Yuen K, Yung AR. Ultra high risk (UHR) for psychosis criteria: are
integrative neuroimaging studies in schizophrenia targeting for early interven- there different levels of risk for transition to psychosis? Schizophr Res
tion and prevention (IN-STEP) project. Schizophr Res 2013;143(1):116–24. 2011;125(1):62–8.
[53] Kotlicka-Antczak M, Pawełczyk T, Rabe-Jabłońska J, Pawełczyk A. PORT [78] Nelson B, Thompson A, Yung AR. Basic self-disturbance predicts psychosis
(Programme of Recognition and Therapy): the first Polish recognition and onset in the ultra high risk for psychosis ‘‘prodromal’’ population. Schizophr
treatment programme for patients with an at-risk mental state. Early Interv Bull 2012;38(6):1277–87.
Psychiatry 2014. http://dx.doi.org/10.1111/eip.12146 [Epub ahead of print]. [79] Nelson B, Yuen HP, Wood SJ, Lin A, Spiliotacopoulos D, Bruxner A, et al. Long-
[54] Koutsouleris N, Riecher-Rössler A, Meisenzahl EM, Smieskova R, Studerus E, term follow-up of a group at ultra high risk (‘‘prodromal’’) for psychosis: the
Kambeitz-Ilankovic L, et al. Detecting the psychosis prodrome across high- PACE 400 study. JAMA Psychiatry 2013;70(8):793–802.
risk populations using neuroanatomical biomarkers. Schizophr Bull 2014. [80] Nieman DH, Ruhrmann S, Dragt S, Soen F, van Tricht MJ, Koelman JH, et al. Psy-
http://dx.doi.org/10.1093/schbul/sbu078 [Epub ahead of print]. chosis prediction: stratification of risk estimation with information-processing
and premorbid functioning variables. Schizophr Bull 40(6):1482–90, doi:10.1093/ Besonderheiten ausreichend berücksichtigt ? Z Kinder Jugendpsychiatr Psy-
schbul/sbt145. chother 2011;39(5):301–11.
[81] Olesen J, Gustavsson A, Svensson M, Wittchen HU, Jönsson B, CDBE2010 study [108] Schultze-Lutter F, Ruhrmann S, Fusar-Poli P, Bechdolf A, Schimmelmann BG,
group. et al. The economic cost of brain disorders in Europe. Eur J Neurol Klosterkötter J. Basic symptoms and the prediction of first-episode psychosis.
2012;19(1):155–62. Curr Pharm Des 2012;18(4):351–7.
[82] Overall JE, Gorham DR. The brief psychiatric rating scale. Psychol Rep [109] Schultze-Lutter F, Marshall M, Koch E. Schizophrenia proneness instrument,
1962;10:799–812. child and youth version; extended English translation (SPI-CY EET). Rome:
[83] Perälä J, Suvisaari J, Saarni SI, et al. Lifetime prevalence of psychotic and bipolar I Giovanni Fioriti Editore s.r.l; 2012.
disorders in a general population. Arch Gen Psychiatry 2007;64(1):19–28. [110] Schultze-Lutter F, Schimmelmann BG, Ruhrmann S, Michel C. ‘A rose is a rose
[84] Phillips LJ, Yung AR, McGorry PD. Identification of young people at risk of is a rose’, but at-risk criteria differ. Psychopathology 2013;46(2):75–87.
psychosis: validation of Personal Assessment and Crisis Evaluation Clinic [111] Schultze-Lutter F, Michel C, Schimmelmann BG. Prevalence of at-risk criteria of
intake criteria. Aust N Z J Psychiatry 2000;34(Suppl.):164–9. psychosis in children and adolescents, and in young adults: results from two
[85] Raballo A, Meneghelli A, Cocchi A, Sisti D, Rocchi MB, Alpi A, et al. Shades of Swiss community samples. Schizophr Res 2014;153(Suppl. 1):134 [Abstract].
vulnerability: latent structures of clinical caseness in prodromal and early [112] Schultze-Lutter F, Klosterkötter J, Ruhrmann S. Improving the clinical pre-
phases of schizophrenia. Eur Arch Psychiatry Clin Neurosci 2014;264(2): diction of psychosis by combining ultra-high risk criteria and cognitive basic
155–69. symptoms. Schizophr Res 2014;154(1–3):100–6.
[86] Riecher-Rössler A, Aston J, Ventura J, Merlo M, Borgwardt S, Gschwandtner U, [113] Simon AE, Grädel M, Cattapan-Ludewig K, Gruber K, Ballinari P, Roth B, et al.
et al. [The Basel Screening Instrument for Psychosis (BSIP): development, Cognitive functioning in at-risk mental states for psychosis and 2-year
structure, reliability and validity]. Fortschr Neurol Psychiatr 2008;76(4): clinical outcome. Schizophr Res 2012;142(1–3):108–15.
207–16. [114] Thorup A, Waltoft BL, Pedersen CB, Mortensen PB, Nordentoft M. Young
[87] Riecher-Rössler A, Pflueger MO, Aston J, Borgwardt SJ, Brewer WJ, males have a higher risk of developing schizophrenia: a Danish register
Gschwandtner U, et al. Efficacy of using cognitive status in predicting study. Psychol Med 2007;37(4):479–84.
psychosis: a 7-year follow-up. Biol Psychiatry 2009;66(11):1023–30. [115] Tolbert HA. Psychoses in children and adolescents: a review. J Clin Psychiatry
[88] Ross RG, Heinlein S, Tregellas H. High rates of comorbidity are found in 1996;57(Suppl. 3):4–8 [discussion 46–7].
childhood-onset schizophrenia. Schizophr Res 2006;88(1–3):90–5. [116] van der Gaag M, Nieman DH, Rietdijk J, Dragt S, Ising HK, Klaassen RM, et al.
[89] Rössler W, Salize HJ, van Os J, Riecher-Rössler A. Size of burden of schizo- Cognitive behavioral therapy for subjects at ultrahigh risk for developing
phrenia and psychotic disorders. Eur Neuropsychopharmacol 2005;15(4): psychosis: a randomized controlled clinical trial. Schizophr Bull 2012;38(6):
399–409. 1180–8.
[90] Rubio JM, Sanjuán J, Flórez-Salamanca L, Cuesta MJ. Examining the course of [117] van Os J, Kapur S. Schizophrenia. Lancet 2009;374(9690):635–45.
hallucinatory experiences in children and adolescents: a systematic review. [118] van Os J, Kenis G, Rutten BP. The environment and schizophrenia. Nature
Schizophr Res 2012;138(2–3):248–54. 2010;468(7321):203–12.
[91] Ruhrmann S, Bechdolf A, Kühn KU, Wagner M, Schultze-Lutter F, Janssen B, [119] Velthorst E, Nieman DH, Becker HE, van de Fliert R, Dingemans PM, Klaassen
et al. Acute effects of treatment for prodromal symptoms for people puta- R, et al. Baseline differences in clinical symptomatology between ultra high
tively in a late initial prodromal state of psychosis. Br J Psychiatry 2007;51 risk subjects with and without a transition to psychosis. Schizophr Res
(Suppl.):88–95. 2009;109(1–3):60–5.
[92] Ruhrmann S, Schultze-Lutter F, Salokangas RK, Heinimaa M, Linszen D, [120] Velthorst E, Derks EM, Schothorst P, Becker H, Durston S, Ziermans T, et al.
Dingemans P, et al. Prediction of psychosis in adolescents and young adults Quantitative and qualitative symptomatic differences in individuals at Ultra-
at high risk: results from the prospective European prediction of psychosis High Risk for psychosis and healthy controls. Psychiatry Res 2013;210(2):
study. Arch Gen Psychiatry 2010;67(3):241–51. 432–7.
[93] Salokangas RK, Patterson P, Heinimaa M, Svirskis T, From T, Vaskelainen L, [121] Velthorst E, Nelson B, Wiltink S, de Haan L, Wood SJ, Lin A, et al. Transition to
et al. Perceived negative attitude of others predicts transition to psychosis in first episode psychosis in ultra high risk populations: does baseline func-
patients at risk of psychosis. Eur Psychiatry 2012;27(4):264–6. tioning hold the key? Schizophr Res 2013;143(1):132–7.
[94] Schaeffer JL, Ross RG. Childhood-onset schizophrenia: premorbid and pro- [122] Walder DJ, Mittal V, Trotman HD, McMillan AL, Walker EF. Neurocognition
dromal diagnostic and treatment histories. J Am Acad Child Adolesc Psychi- and conversion to psychosis in adolescents at high-risk. Schizophr Res
atry 2002;41(5):538–45. 2008;101(1–3):161–8.
[95] Schimmelmann BG, Schultze-Lutter F. Early detection and intervention of [123] Welsh P, Tiffin PA. The ‘at-risk mental state’ for psychosis in adolescents:
psychosis in children and adolescents: urgent need for studies. Eur Child clinical presentation, transition and remission. Child Psychiatry Hum Dev
Adolesc Psychiatry 2012;21(5):239–41. 2014;45(1):90–8.
[96] Schimmelmann BG, Conus P, Cotton S, McGorry PD, Lambert M. Pre-treat- [124] Whithing J, Rutjes AW, Dinnes J, Reitsma BJ, Bossuyt PM, Kleijnen J. Devel-
ment, baseline, and outcome differences between early-onset and adult- opment and validation of methods assessing the quality of diagnostic accu-
onset psychosis in an epidemiological cohort of 636 first-episode patients. racy studies. Health Tech Assess 2004;8(25):1–234 [iii].
Schizophr Res 2007;95(1–3):1–8. [125] Wittchen HU, Jacobi F, Rehm J, Gustavsson A, Svensson M, Jönsson B, et al. The
[97] Schimmelmann BG, Huber CG, Lambert M, Cotton S, McGorry PD, Conus P. size and burden of mental disorders and other disorders of the brain in
Impact of duration of untreated psychosis on pre-treatment, baseline, and Europe 2010. Eur Neuropsychopharmacol 2011;21(9):655–79.
outcome characteristics in an epidemiological first-episode psychosis cohort. [126] Woodberry KA, Seidman LJ, Giuliano AJ, Verdi MB, Cook WL, McFarlane WR.
J Psychiatr Res 2008;42(12):982–90. Neuropsychological profiles in individuals at clinical high risk for psychosis:
[98] Schimmelmann BG, Schmidt SJ, Carbon M, Correll CU. Treatment of adolescents relationship to psychosis and intelligence. Schizophr Res 2010;123(2–3):188–98.
with early-onset schizophrenia spectrum disorders: in search of a rational, [127] World Health Organization. Prevention of mental disorders: effective inter-
evidence-informed approach. Curr Opin Psychiatry 2013;26(2): 219–30. ventions and policy options. Geneva: World Health Organization; 2004.
[99] Schimmelmann BG, Walger P, Schultze-Lutter F. The significance of at-risk [128] Yung AR, McGorry PD. The initial prodrome in psychosis: descriptive and
symptoms for psychosis in children and adolescents. Can J Psychiatry qualitative aspects. Aust N Z J Psychiatry 1996;30(5):587–99.
2013;58(1):32–40. [129] Yung AR, Phillips LJ, Yuen HP, Francey SM, McFarlane CA, Hallgren M, et al.
[100] Schlosser DA, Jacobson S, Chen Q, Sugar CA, Niendam TA, Li G, et al. Recovery Psychosis prediction: 12-month follow up of a high-risk (‘‘prodromal’’)
from an at-risk state: clinical and functional outcomes of putatively prodro- group. Schizophr Res 2003;60(1):21–32.
mal youth who do not develop psychosis. Schizophr Bull 2012;38(6): [130] Yung AR, Yuen HP, McGorry PD, Phillips LJ, Kelly D, Dell’Olio M, et al. Mapping
1225–33. the onset of psychosis: the comprehensive assessment of at-risk mental
[101] Schmidt SJ, Grunert VM, Schimmelmann BG, Schultze-Lutter F, Michel C. states. Aust N Z J Psychiatry 2005;39(11–12):964–71.
Differences in coping, self-efficacy, and external control beliefs between [131] Yung AR, Phillips LJ, Simmons MB, Ward J, Thompson P, French P, et al.
patients at-risk for psychosis and patients with first-episode psychosis. CAARMS. Comprehensive assessment at risk mental states. Parkville Victoria:
Psychiatry Res 2014;219(1):95–102. The PACE Clinic, ORYGEN Research Centre, University of Melbourne, Depart-
[102] Schultze-Lutter F. Subjective symptoms of schizophrenia in research and the ment of Psychiatry; 2006.
clinic: the basic symptom concept. Schizophr Bull 2009;35(1):5–8. [132] Yung AR, Stanford C, Cosgrave E, Killackey E, Phillips L, Nelson B, et al. Testing
[103] Schultze-Lutter F, Ruhrmann S, Klosterkötter J. In: Johannessen JO, Martin- the ultra high risk (prodromal) criteria for the prediction of psychosis in a
dale B, Cullberg J, editors. Evolving psychosis. Different stages, different clinical sample of young people. Schizophr Res 2006;84(1):57–66.
treatments. London, New York: Routledge; 2006. p. 104–23. [133] Yung AR, Nelson B, Stanford C, Simmons MB, Cosgrave EM, Killackey E, et al.
[104] Schultze-Lutter F, Addington J, Ruhrmann S, Klosterkötter J. Schizophrenia Validation of ‘‘prodromal’’ criteria to detect individuals at ultra high risk of
proneness instrument adult version (SPI-A). Rome: Giovanni Fioriti Editore psychosis: 2-year follow-up. Schizophr Res 2008;105(1–3):10–7.
s.r.l; 2007. [134] Zhang T, Li H, Woodberry KA, Seidman LJ, Zheng L, Li H, et al. Prodromal
[105] Schultze-Lutter F, Klosterkötter J, Picker H, Steinmeyer EM, Ruhrmann S. psychosis detection in a counseling center population in China: an epidemi-
Predicting first-episode psychosis by basic symptom criteria. Clin Neuropsy- ological and clinical study. Schizophr Res 2014;152(2–3):391–9.
chiatry 2007;4(1):11–22. [135] Ziermans TB, Schothorst PF, Sprong M, van Engeland H. Transition and
[106] Schultze-Lutter F, Ruhrmann S, Berning J, Maier W, Klosterkötter J. Basic remission in adolescents at ultra-high risk for psychosis. Schizophr Res
symptoms and ultrahigh risk criteria: symptom development in the initial 2011;126(1–3):58–64.
prodromal state. Schizophr Bull 2010;36(1):182–91. [136] Zimmermann R, Gschwandtner U, Wilhelm FH, Pflueger MO, Riecher-Rössler
[107] Schultze-Lutter F, Resch F, Koch E, Schimmelmann BG. Früherkennung von A, Fuhr P. EEG spectral power and negative symptoms in at-risk individuals
Psychosen bei Kindern und Adoleszenten – sind entwicklungsbezogene predict transition to psychosis. Schizophr Res 2010;123(2–3):208–16.
Supplementary Material to:
EPA guidance on the early detection of clinical high risk states of psychoses
Frauke Schultze-Lutter, Chantal Michel, Stefanie J. Schmidt, Benno G. Schimmelmann,
Nadja P. Maric, Raimo R.K. Salokangas, Anita Riecher-Rössler, Mark van der Gaag, Merete
Nordentoft, Andrea Raballo, Anna Meneghelli, Max Marshall, Anthony Morrison, Stephan
Ruhrmann, Joachim Klosterkötter
Contents
S.1. Procedures and formulae of meta-analyses ..................................................................... 2

S.2. Sensitivity analyses ........................................................................................................... 3
S.3. Supplementary Table 1 ..................................................................................................... 4
S.4. Supplementary Table 2 ................................................................................................... 24
S.8. References of Supplementary Material ........................................................................... 32
Supplementary Material: EPA guidance on the early detection of CHR states of psychoses 2/36
S.1. Procedures and formulae of meta-analyses
Although proportions are strictly speaking no measure of effect, they are commonly used in
meta-analyses of univariate studies when the assessment of measure of interest is
comparable across studies [1]. As this can be assumed in early detection studies in that first-
episode psychosis is generally diagnosed according to DSM-IV, we used the proportions of
𝑘𝑖
conversions at follow-up (Ei = with ki = number of patients having developed psychosis at
𝑛𝑖
𝑘𝑖 𝑘𝑖
follow-up tx, and ni = (sub)sample size at baseline) and their variance (Vi = (1 – ) / ni) as
𝑛𝑖 𝑛𝑖
effect estimates in a fixed-effects model [1]. The inverse variance was used as weight i to
account for the different sample size of studies [1]. Pooled effects and their variance were
∑𝑘
𝑖=1 𝜔𝑖×𝐸𝑖 ∑𝑘
𝑖=1 𝜔𝑖
calculated as 𝐸 = ∑𝑘
and 𝑉 = ; 95% confidence intervals (CIs) of single and
𝑖=1 𝜔𝑖 (∑𝑘
𝑖=1 𝜔𝑖)²
pooled effects were calculated as Ei  (0.98Vi). Pooled sample effects were tested by the z-
𝐸̅
statistic with z = ̅
and were assumed significant at  = 5% when z  1.96 and at  = 1%
√𝑉
when z  2.58 [1].
Heterogeneity between Eis included in 𝐸 were tested by the Q-statistic, a type of 2-statistic
(𝐸𝑖− 𝐸̅ )²
with df = l – 1 and l = number of Eis [1]. The formula used was Q = ∑𝑙𝑖=1 . Additionally,
𝑉𝑖
𝑄−𝑑𝑓
I2 (= × 100%) was calculated as an estimate of the relative size of heterogeneity [1]. In
𝑄
line with Higgins et al. [2], I2 values of 25%, 50% and 75% were regarded as signifying low,
moderate and high heterogeneity; and negative values of I2 were put at zero.
Whenever significant heterogeneity was detected indicating that considerable variance might
have been introduced by sources other than the sampling error considered in the fixed-
effects model, a random-effects model was applied [1]. Thereby, the additional test variance
𝑄−𝑑𝑓
2 was calculated as 2 = ∑𝑙𝑖=1 𝜔𝑖−(∑𝑙𝑖=1 𝜔𝑖² / ∑𝑙𝑖=1 𝜔𝑖)
and added to Vi to calculate the variance of
Ei (Vi*); if 2 takes on a negative value, this is usually interpreted as meaning that the random-
effects variance is inconsequential (i.e., equal to zero) and the random-effects model
collapses to a fixed-effects model meta-analysis [3]. The inverse value of Vi* provided the
weight i*. Using Vi* and i* in the same formulae as in the fixed-effect model, 𝐸 ∗ , 𝑉 ∗ as well
as the related 95% CIs and z-values were computed.
S.2. Sensitivity analyses
To estimate the influence of assessment scales and, relatedly, definitions of UHR criteria
(SIPS, CAARMS early versions, CAARMS 2006 version), type of ARMS criteria and
combinations (APS, BLIPS, GRFD, COPER, COGDIS, UHR plus COGDIS, UHR and/or
COGDIS), and age characteristic of the sample (CAD, YOUTH or ADULT), the above
analyses (see S.1.) were performed with these subgroups in addition.
The resulting pooled effect sizes that are essentially proportions were converted into
percentages of conversion rates (𝐸 / 𝐸 ∗  100%) and compared for significant differences
using exploratory one-dimensional 2-tests (with df = number of effect sizes – 1) unadjusted
for multiple testing.

S.3. Supplementary Table 1
STable 1 List and description of studies included in the meta-analysis (main text reference number in blue) and the guidance authors’ rating of
the grade of evidence (GE)
Study, Setting & response/ Sample characteristics Assessment and Follow-up & missing Annual conversion rate Conversion assessment &
GE participation rate (size, age, gender, co- baseline distribution of observations diagnoses
morbidities) CHR criteria (cumulative)
Australia
[77] Specialized early N=311 BPRS (n=407-409) and 14.9 yrs. Overall: n=114 (36.7%) Conversion according to
Nelson et al. detection service: CAARMS (before 2006 (min. 2.4 yrs.; 7.53.2 BPRS/CAARMS or state public
2013 [4] Personal Assessment Age: 14-30 yrs. version; n=389-397) yrs.) 1 yr.: n=65 (20.9%) mental health records
and Crisis Evaluation (Mdn=18 yrs.) 2 yrs.: n=79 (25.4%) Note: Converters might not have
PACE 400 (PACE) clinic Of eligible sample: Missing observations: 3 yrs.: n=94 (30.2%) sought help after conversion and,
study Assigned age group: YOUTH APS: n=316 (79.4%) 2-4 yrs.: n=0 (0%) 4 yrs.: n=102 (32.8%) consequently, might not show in
Mixed: includes 3 BLIPS: n=56 (14.1%) 4-6 yrs.: n=52 state public mental health records
GE: 2+ intervention and 4 % male: 48.1% GRFD: n=115 (28.9%) (16.7%) Estimated conversion rates
observational studies 6-8 yrs.: n=135 (Kaplan-Meier) Diagnoses: Not reported
conducted between Co-morbidities: Not reported (43.4%) 1 yr.: 16.5% (12.7-20.1)
1993 and 2006. 8-10 yrs.: n=179 2 yrs.: 20.4% (16.3-24.4)
(57.5%) 3 yrs.: 24.9% (20.4-29.2)
% response: 74.8% of 10-12 yrs.: n=214 4 yrs.: 27.6% (22.8-32.1)
eligible sample (68.8%) 5 yrs.: 30.1% (25.0-34.8)
(n=416) 12-15 yrs: n=285 10-15 yrs.: 34.9% (28.7-
(91.6%) 40.6)
[78] Specialized early N=817 CAARMS (before 2006 6 mths. Overall: n=72 (8.8%) Conversion according to CAARMS
Nelson et al. detection service: version) or state public mental health
2011 [5] PACE clinic Age: 14-29 yrs. (Mdn: 18 yrs.) Missing observations: APS: n=62 (9.3%) records
APS: n=664 (81.3%) for conversion: 0% BLIPS: n=5 (13.9%) Note: Converters might not have
Partly includes Mixed: also includes Assigned age group: YOUTH BLIPS: n=36 (4.4%) CAARMS follow-up GRFD: n=13 (6.2%) sought help after conversion and,
PACE 400 participants of GRFD: n=209 (25.6%) assessment: n=307 APS+GRFD: n=8 (8.7%) consequently, might not show in
sample intervention studies % male: 40.8% APS+GRFD: n=92 (37.6%) state public mental health records
(n=208) but (11.3%)
GE: 2+ predominately Co-morbidities: Not reported Schizophrenia spectrum disorder:
observational study 23%
on patients presented Psychotic Disorder NOS: 53%
between 01/2000 and All non-affective psychosis: 76%
11/2008
Mood disorder with psychotic
% response: 88.0% of features: 14%
eligible sample All affective psychosis: 14%
(n=928)
GE participation rate (size, age, gender, co- baseline distribution of observations per diagnoses
morbidities) CHR criteria follow-up
(cumulative)
[79] Specialized early N=49 CAARMS 2006 version 569±345 days (Mdn: Overall: n=13 (26.5%) Conversion according to CAARMS
Nelson et al. detection service: 676 days) or state public mental health
2012 [6] PACE clinic Age: 19.2±2.9 yrs. (15-25 yrs.) APS: n=37 (75.5%) Cumulative conversion rate records
BLIPS: n=1 (2.0%) Missing observations: in % ±SE (Kaplan-Meier) Note: Converters might not have
GE: 2+ Observational study Assigned age group: MIX GRFD: n=4 (8.2%) for conversion: 0% 6 mths.: 22.8±4.0 sought help after conversion and,
of patients presenting APS+GRFD: n=7 CAARMS follow-up 1 yr.: 24.9±3.6 consequently, might not show in
between 05/2008 and % male: 44.9% (14.3%) assessment: n=8 2 yrs.: 27.6±3.7 state public mental health records
07/2010, includes (16.3%)
some of [15] Co-morbidities: 57.1% with Schizophrenia spectrum disorder:
mood disorder, 16.3% with 61.5%
% response: Not anxiety disorder, 8.2% with
reported other axis-I disorder; 14.3% Other psychotic diagnosis (incl.
schizotypal personality affective psychosis and psychosis
disorder NOS): 38.5%
[80,81] No specialized early N=292 CAARMS (before 2006 2 yrs. UHR: Conversion according to CAARMS
Yung et al. detection service: version) 6 mths.: n=12 (10.1%) or state public mental health
2006, 2008 ORYGEN Youth Age: mean: 18.1 yrs. (15-24 Missing observations: 2 yrs.: n=19 (15.9%) records
[7,8] Health yrs.) UHR: n=119 (40.7%) for conversion: 0% Note: Converters might not have
APS: n=111 (38.0%) CAARMS 6-mths. none: sought help after conversion and,
GE: 2+ Observational study Assigned age group: YOUTH BLIPS: 0% assessment: n=97 6 mths.: n=1 (0.6%) consequently, might not show in
including some PACE GRFD: n=13 (4.5%) (33.2%) 2 yrs.: n=2 (1.2%) state public mental health records
400 participants of 04- % male: 48.9% none: n=173 (59.2%) CAARMS 2-yrs.
10/2003 (n76) assessment: n=99 Diagnoses: Not reported
Co-morbidities: Only reported (33.9%)
% response: 76.6% of for Youthscope subsample
eligible sample (n=149): 23.4% any axis-I
(n=381) disorder; 46.3% mood
disorders, 42.3% anxiety
disorders, 22.1% substance
use disorders, 7.4% eating
disorders
(cumulative)
[82] Specialized early N=92 CAARMS 2006 version 26 mths. Overall: n=20 (21.7%) Conversion according to CAARMS
Bechdolf et al. detection service: (min. 14 mths.,
2010 [9] PACE clinic Age: 18.0±3.0 yrs. (15-24 yrs.) APS: n=74 (80.4%) 682±283 days) Schizophrenia: 30%
BLIPS: n=5 (5.4%) Delusional disorder: 5%
GE: 2+ Observational study Assigned age group: YOUTH GRFD: n=28 (30.4%) Missing observations: Unspecified acute psychotic
of PACE patients Not reported disorder: 50%
presenting in 2007, % male: 34.8% All non-affective psychosis: 85%
includes some of [13]
Co-morbidities: 56.5% Depression with psychotic features:
% response: Not depression, 8.7% anxiety 15%
reported disorder, 8.7% dysthymia or All affective psychosis: 15%
cyclothymia, 15.2% PTSD,
3,3% adjustment disorder
[67] Specialized early N=74 CAARMS (before 2006 1 yr. (26.3±9.2 Overall: n=37 (50.0%) Conversion according to DSM-IV
Mason et al. detection service: version) mths.) and/or CAARMS
2004 [10] Psychological Age: 17.3±2.8 yrs. (13-28 yrs.) APS: n= 22 (51.2%)
Assistance Service APS: n=43 (58.1%) Missing observations: BLIPS: n=14 (60.8%) Schizophrenia: 18.9%
GE: 2+ (PAS) of Hunter Assigned age group: YOUTH BLIPS: n=23 (31.1%) Not reported GRFD: n=2 (10.5%) Schizoaffective disorder: 27%
Mental Health in New GRFD: n=19 (25.7%)
South Wales, % male: 52.7% Depression with psychotic features:
Australia 18.9%
Co-morbidities: Not reported Mania with psychotic features:
Observational study 5.4%
Bipolar disorder with psychotic
% response: 56.9% of features: 5.4%
eligible sample
(n=130) Unspecified psychotic episode
according to CAARMS: 24.3%
(cumulative)
[83] Specialized early N=30 CAARMS 2006 version 2 yrs. 6 mths: n=1 (3.4%) Conversion according to CAARMS
Welsh & Tiffin detection service: 1 yr.: n=1 (3.4%) and/or medical records
2013 [11] Follow-up of At-Risk Age: 15.8±1.4 yrs. (12-18 yrs.) APS: n=30 (100%) Missing observations: 2 yrs.: n=2 (7.1%) Note: Converters might not have
Mental State for GRFD: n=4 (13.3%) 6 mths.: n=1 (3%) sought help after conversion and,
GE: 2+ Psychosis – FARMS Assigned age group: CAD 1 yr.: n=4 (13%) consequently, might not show in
Clinic 2 yrs.: n=2 (6.7%) state public mental health records
% male: 47%
Observational study CAARMS 1- and 2- Schizophrenia: 50%
Co-morbidities: 43% yrs. assessments: Other psychotic disorder: 50%
% response: Not depressive disorders, 20% majority All non-affective psychosis:
reported anxiety disorder, 17% 100%
pervasive developmental
disorder, 7% behavioural
disorders, 7% other disorders
North America
[70] Mixed, centres with N=291 SIPS 3.0 2.5 yrs. Overall: n=82 (28.2%) Conversion according to SIPS
Cannon et al. and without (non-converters:
2008 [12] ; specialized early Age: 18.1±4.6 yrs. (12-30 yrs.) APS: n=282 (96.9%) 575±258 days) APS: n=79 (28.0%) Diagnoses: Not reported
Addington et detection services BLIPS: n=7 (2.4%) BLIPS: n=3 (42.9%)
al., 2007 [13] included. Assigned age group: YOUTH GRFD: n=2 (0.7%) Missing observations: GRFD: 0%
Not reported
North Pooled sample from % male: 58.4% Cumulative conversion rate
American initially independent 7 in % ±SE (Kaplan-Meier)
Prodrome observational and Co-morbidities of eligible 6 mths.: 12.7±1.9
Longitudinal intervention studies sample (n=370): 34.9% with 1 yr.: 21.7±2.5
Study with 1 follow-up mood disorder, 30.4% with 2 yrs.: 32.6±3.3
(NAPLS 1) anxiety disorder, 15.3% with 2.5 yrs.: 35.3±3.7
% response: 78.6% of alcohol abuse or dependence,
GE: 2+ eligible sample 19.5% with drug 1 yr.: n=63 (21.7%)
(n=370) abuse/dependence 2 yrs.: n=95 (32.6%)
2.5 yrs.: n=103 (35.3%)
(cumulative)
[71] Specialized early N=125 SIPS 3.0 2 yrs. Overall: n=27 (21.6%) Conversion according to SIPS
Schlosser et detection service:
al. 2012 [14] Staglin Music Festival Of n=84 with 1 follow-up: Of n=84 with 1 follow- Missing observations: Of n=84 with 1 follow-up: Diagnoses: Not reported
Centre for the Age: 16.9±3.5 yrs. up: overall: n=41 (32.8%) APS: n=17 (26.2%)
GE: 2+ Assessment and APS: n=65 (78.4%) no follow-up BLIPS: n=10 (58.8%)
Prevention of Assigned age group: YOUTH BLIPS: n=17 (20.2%) assessment GRFD: n=0 (0%)
Prodromal States GRFD: n=2 (2.4%)
(CAPP) % male: 62%
Observational study Co-morbidities: 47.6% with

anxiety disorder, 44% with
% response: Not mood disorder
reported
[72] Specialized early N=101 SIPS 3.0 Mean: 3.0±1.6 yrs. Overall: n=15 (14.9%) Conversion according to SIPS
Carrión et al. detection service: (Mdn=2.8 yrs.)
2013 [15] Recognition and Of n=92 with 1 follow-up: Of n=92 with 1 follow- Diagnoses: Not reported
Prevention (RAP) Age: 15.9±2.2 yrs. (12-22 yrs.) up: Missing observations:
GE: 2+ program, Glen Oaks, APS: n=92 (100%); Overall: n=9 (9%) no
NY Assigned age group: CAD follow-up assessment
BLIPS as exclusion
Observational study, % male: 63.0% criterion
some participants
included in NAPLS 1 Co-morbidities: 63% mood
disorders, 57.6% anxiety
% response: Not disorders, 9.8% substance-
reported use disorders
[73] Specialized early N=73 SIPS 2 yrs. APS and GRFD only Conversion defined as
Woodberry et detection service: (n=68): development of any positive item
al. 2010 [16] Portland Identification Age: 16.5±2.5 yrs. (12-25 yrs.) APS: n=65 (89%) Missing observations: n=13 (19.1%) rated 6 on the SIPS
and Early Referral BLIPS: n=5 (7%) 2 yrs.: n=16 (21.9%);
GE: 2+ (PIER) program Assigned age group: YOUTH GRFD: n=3 (4%) only mean follow-up Diagnoses: Not reported
of 7 mths.
Observational study % male: 53% BLIPS excluded from
analyses, because
% response: 81.1% Co-morbidities: Not reported included in conversion
of eligible sample criteria
(n=90)
(cumulative)
[74] Specialized early N=24 of supportive therapy SIPS 3.0 1.5 yrs. Overall: n=3 (12.5%) Conversion according to SIPS
Addington et detection service: condition
al. 2011 [17] PRIME Clinic at the APS: n=24 (100%) Missing observations: 6 mths.: n=3 (12.5%) Schizophrenia: 100%
Centre for Addiction Age: 21.1±3.74 yrs. (14-30 BLIPS: 0% 6 mths.: n=8 (33.3%) 1yr.: n=3 (12.5%) All non-affective psychosis:
GE: 2+ and Mental Health, yrs.) GRFD: 0% 1 yr.: n=9 (37.5%) 2 yrs.: n=3 (12.5%) 100%
Toronto 1.5 yrs.: n=11
Assigned age group: ADULT (45.8%)
Intervention study:
CBT vs. supportive % male: 35.3%
therapy
Co-morbidities of all
% response: 50.0% of randomized participants
eligible sample (n=51): 25.5% mood
(n=112) consented to disorders, 17.7% anxiety
the study (n=56), 5 disorders, 5.9% alcohol abuse,
dropped out before 9.8% cannabis abuse
randomization
[75] Specialized early N=170 SIPS 5.0 4 yrs. Overall: n=29 (17.1%) Conversion according to SIPS
Buchy et al. detection services:
2014 [18] PRIME clinics of the Age: 19.7±4.5 yrs. (12-31 yrs.) APS: n=167 (98.2%) Missing observations: Diagnoses: Not reported
Universities of BLIPS: 0% Not reported
Enhancing the Toronto, North Assigned age group: MIX GRFD: n=6 (3.5%)
Prospective Carolina and Yale
Prediction % male: 56.5%
Psychosis Observational study
(PREDICT) Co-morbidities: Not reported
study % response: Not
reported
GE: 2+
(cumulative)
[76] Specialized early N=24 SIPS 3.0 4 yrs. 4 yrs.: n=3 (12.5%) Conversion according to SIPS
Kayser et al. detection service:
2013 [19] Centre of Prevention Of n=21 analysed: APS: n=21 (100%) Missing observations: Diagnoses: “typically
and Evaluation Age: 21.4±3.8 yrs. (13-27 yrs.) BLIPS: 0% Overall: n=3 (12.5%) schizophrenia”
GE: 2+ (COPE), New York GRFD: 0%
State Psychiatric Assigned age group: ADULT All conversions in patients of age
Institute, Columbia 16 years
University % male: 48.1%
Observational study Co-morbidities: Not reported
% response: Not
reported
Germany
[84,85] No specialized early N=160 BSABS 9.6±7.6 yrs. Of those with the criterion: Conversion to schizophrenia
Klosterkötter detection service. (5-37 yrs.) COPER according to DSM-IV
et al. 2001 Outpatient Age: 29.3±10.0 yrs. (15-53 Intake criteria: clinical 1 yr.: n=21 (19.8%)
[20]; Schultze- departments of yrs.) suspicion of beginning Missing observations: 2 yrs.: n=39 (36.8%) Schizophrenia: 100%
Lutter et al. German psychiatric psychosis and >4 years: 0% 3 yrs.: n=53 (50.0%) All non-affective psychosis:
2006 [21] university Assigned age group: ADULT assessment for basic >3 yrs.: n=69 (65.1%) 100%
departments symptoms, no past or
Cologne Early % male: 52.5% present psychosis COGDIS
Recognition Observational study 1 yr.: n=16 (23.9%)
(CER) study Co-morbidities: 36.2% COPER: n=106 (66.3%) 2 yrs.: n=31 (46.3%)
on that % response: 42% of personality disorders, 29.4% COGDIS: n=67 (41.88%) 3 yrs.: n=41 (61.2%)
COPER and eligible sample affective disorders, 17.5% none: n=54 (33.8%) >3 yrs.: n=53 (79.1%)
COGDIS were (n=385) somatoform disorders, 16.9%
developed anxiety disorders
GE: 2+
(cumulative)
[86] Specialized early N=246 SPI-A and 4 yrs. (min. 1 yr.) 4 yrs. total sample: n=81 Conversion according to DSM-IV
Schultze- detection service: SIPS 3.0 (32.9%) using SCID-I
Lutter et al. FETZ Cologne Age: 24.9±6.0 yrs. (15-39 Missing observations: 4 yrs. CHR sample: n=75
2014 [22] yrs.); COGDIS: n=157 (63.8%) 1 yr.: 0% (38.7%) Schizophrenia: 75.3%
Observational study 18% minors (7.3%) 2 yrs.: n=11 (4.5%) Schizophreniform disorder: 3.7%
GE: 2+ APS: n=157 (63.8%) 3 yrs.: n=31 (12.6%) Annual hazard rates: Schizoaffective disorder: 1.2%
% response: 51% of Assigned age group: ADULT BIPS: n=22 (8.9%) 4 yrs.: n=57 (23.2%) 1 yr. only COGDIS: 0.11 Delusional disorder: 6.2%
initial sample of GRFD: n=0 (0%) 1 yr. only UHR: 0.28 Substance-induced psychosis:
n=482 male: 63.0% none: n=52 (21.1%) 1 yr. UHR+COGDIS: 0.36 4.9%
2 yrs. only COGDIS: 0.14 All non-affective psychosis:
Co-morbidities: 64.6% any 2 yrs. only UHR: 0.28 91.4%
axis-I disorder incl. 30.9% 2 yrs. UHR+COGDIS: 0.53
depressive disorders 3 yrs. only COGDIS: 0.23 Depression with psychotic features:
3 yrs. only UHR: 0.28 3.7%
3 yrs. UHR+COGDIS: 0.61 Bipolar disorder with psychotic
4 yrs. only COGDIS: 0.23 features: 4.9%
4 yrs. only UHR: 0.28 All affective psychosis: 8.6%
4 yrs. UHR+COGDIS: 0.66
[87] Mixed, centres with N=65 of supportive ERIraos 2 yrs. 1 yr.: n=9 (13.8%) Conversion assessed with PANSS:
Bechdolf et al. and without counselling condition 2 yrs.: n=10 (15.4%) any positive psychotic symptom in
2012 [23] specialized early COPER: n=64 (98.5%) Missing observations: psychotic intensity for >7days
detection services Age: 26.8±6.2yrs (18-40 yrs.) GRFD: n=21 (32.3%) 1 yr.: n=8 (12.3%) Additional conversion to an
German included. 2 yrs.: n=16 (24.6%) BLIPS/APS defined CHR Schizophrenia/
Research Assigned age group: ADULT Note: APS and BLIPS state: Schizophreniform disorder: 80%
Network on Intervention study: were exclusion criteria at 1 yr.: n=2 (3.1%) Psychosis: 20%
Schizophrenia integrated % male: 64.6% baseline, and, at follow- 2 yrs.: n=3 (4.6%) All non-affective psychosis:
study; psychological up, additional conversion 100%
project 1.1.2 intervention vs. Co-morbidities: Not reported criteria
supportive counselling
GE: 2+
% response: 76% of
eligible sample (128
of 168)
(cumulative)
[88] Specialized early N=45 BSABS and 4 yrs. 4 yrs.: n=14 (31.1%) Conversion according to ICD-10
Koutsouleris et detection service: Age: 25.1±5.8 yrs. (minimum CAARMS (before 2006
al. 2009 [24] Early Detection and age: 18 yrs.) version) Missing observations: Note: annual and single Schizophrenia: 71.4%
Intervention Centre 4 yrs.: n=12 (26.7%) criteria rates below include Schizoaffective disorder: 28.6%
GE: 2+ for Mental Crises, Assigned age group: ADULT COPER and/or GRFD: n=1 conversion to ICD-10 All non-affective psychosis:
Ludwig-Maximilians- n=20 (44.4%) schizotypal disorder and 100%
University % male: 62.2% APS and/or BLIPS: n=25 were therefore not
(55.6%) considered in respective
Observational study Co-morbidities: Not reported moderator analyses
COPER: n=41 (91.1%)
% response: Not APS: n=20 (44.4%) 1 yr.: n=13(2.8%)
reported BLIPS: n=17 (37.8%) 2 yrs.: n=14 (31.1%)
3 yrs.: n=15 (33.3%)
COPER: n=13 (31.7%)

APS: n=12 (60.0%)
BLIPS: n=8 (47.1%)
[89] Specialized early N=146 SPI-A 2 yrs. 2 yrs.: n=48 (32.9%) Conversion assessed with PANSS:
Schultze- detection service: any positive psychotic symptom in
Lutter et al. FETZ Cologne Age: 24.4±5.2 yrs. (16-39 COPER: n=146 (100%) Missing observations: 6 mths.: n=23 (15.8%) psychotic intensity for >7days
2007 [25] yrs.) 6 mths.: n=16 (11%) 1 yr.: n=36 (24.7%)
Observational study 1 yr.: n=24 (16.4%) Schizophrenia: 82.4%
GE: 2+ Assigned age group: ADULT 2 yrs.: n=38 (26.0%) Of additional n=3, Schizophreniform disorder: 7.8%
% response: Not conversion after 2 yrs. Schizoaffective disorder: 3.9%
reported % male: 69.2% became known: Delusional disorder: 3.9%
(overall:n=51 (34.9%)) All non-affective psychosis: 98%
Co-morbidities: Not reported
Depression with psychotic features:
2%
All affective psychosis: 2%
(cumulative)
[90,91] Mixed, centres with N=62 ERIraos 2 yrs. 2 yrs.: n=25 (40.3%) Assessment: SCID-I psychosis
Bodatsch, and without section
Ruhrmann et specialized early Age: 24.8±6.0 yrs. (18-40 yrs.) APS and/or BLIPS: Missing observations:
al. 2011 [26]; detection services 100% only cases included in Schizophrenia: 92%
Ruhrmann et included. Assigned age group: ADULT analyses with Schizophreniform disorder: 4%
al. 2007 [27] conversion within or Delusional disorder: 4%
Observation of the % male: 66.1% follow-up until 2 yrs. All non-affective psychosis:
German control group of an 100%
Research intervention study Co-morbidities: Not reported
Network on
Schizophrenia % response in 2007
study; project [27]: 32.5% (n=124 in
1.1.3 both intervention
groups) of eligible
GE: 2+ sample (n=382)
United Kingdom
[94] Specialized early N=290 CAARMS 2006 and 10 yrs. 4 yrs.: n=44 (15.2%) Conversion: Positive symptom
Fusar-Poli et detection service: SPI-A according to CAARMS for >1 day
al. 2013 [28] Outreach and support Age: 22.9±4.61 yrs. (14-35 Missing observations: Mean time to conversion:
in SOUTH London yrs.) APS: n=258 (89%) Not reported 375 days (95% CIs 280; Diagnoses: Not reported
GE: 2+ (OASIS) BIPS: n=52 (18%) 470 days), last observed
Assigned age group: ADULT GRFD: n=41 (14%) conversion at 1242 days
Observational study (~3.5 yrs)
% male: 56.1% COGDIS (since 2008):
% response: Not not reported
reported Co-morbidities: 14% anxiety
disorders, 29% depressive
disorders, 8% personality
disorders, 13% substance use
disorders, 3% OCD, 5% other
disorders
(cumulative)
[95,96] No specialized early N=23 of TAU condition PANSS 1-3 yrs. 1 yr.: n=6 (26.1%) Conversion according to DSM-IV
Morrison et al. detection service, 3 yrs.: n=7 (30.4%)
2004, 2007 referrals by a variety Age: 21.5±5.2 yrs. (16-36 yrs.) In total sample (n=60) Missing observations: Schizophrenia: 71.4%
[29,30] of clinical institutions APS: n=48 (80%) 1 yr.: n=7 (30.4%) Schizoaffective disorder: 14.3%
Assigned age group: MIX BLIPS: n=6 (10%) 3 yrs.: n=13 (56.5%) Other psychotic DSM-IV disorder:
Early Intervention study: CT GRFD: n=4 (6.7%) 14.3%
Detection and vs. treatment-as-usual % male: 82.6% All non-affective psychosis:
Intervention (TAU) 100%
Evaluation Co-morbidities: Not reported
(EDIE I) trial % response: 95.2% of
eligible sample (n=63)
GE: 2+ was randomized
(n=60)
[97,98] Mixed, centres with N=144 of monitoring condition CAARMS (before 2006 1-2 yrs. 1 yr.: n=10 (6.9%) Conversion according to CAARMS
Morrison et al. and without version) 2 yrs.: n=13 (9.2%) or reports from family doctors
2011, 2012 specialized early Age: 20.8±4.5 yrs. (14-35 yrs.) Missing observations:
[31,32] detection services Only distribution of total 1 yr.: n=51 (35.4%) In total sample (n=23 conversions):
included. Assigned age group: MIX sample (n=288) 2 yrs.: n=65 (45.1%) Schizophrenia: 34.8%
Early APS: n=266 (92.4%) Schizoaffective disorder: 21.7%
Detection and % male: 63.2% BLIPS: n=7 (2.4%) Delusional disorder: 13%
Intervention Intervention study: CT GRFD: n=33 (11.5%) Psychosis NOS: 13%
Evaluation vs. monitoring Co-morbidities of total sample Brief psychotic disorder: 4.3%
(EDIE-2) trial (n=288): 67% 1 DSM-IV All non-affective psychosis:
% response: 100% of diagnosis: 41.3% depressive 100%
GE: 2+ eligible sample of disorders, 19.9% panic
n=288 disorders with / without
agoraphobia, 11.2% social
phobia, 10.9% specific phobia,
8.6% generalized anxiety
disorder, 7.5% OCD, 2.2%
PTSD
(cumulative)
The Netherlands
[57] No specialized early N=72 SIPS 3.0 and 2 yrs. 1 yr. overall: n=7 (9.7%) Conversion according to SIPS
Ziermans et al. detection service: SPI-A (brief) 2 yrs. overall: n=9 (12.5%)
2011 [33] Department of Child Age: 15.3±1.9 yrs. (12-18 yrs.) Missing observations: Schizophrenia: 66.7%
and Adolescent APS: n=65 (90.3%) 1 yr.: n=10 (13.9%) 2 yrs. APS: n=9 (13.8%) Schizoaffective disorder: 11.1%
GE: 2+ Psychiatry, University Assigned age group: CAD BLIPS: n=4 (5.6%) 2 yrs.: n=14 (19.4%) 2 yrs. COGDIS: n=7 Psychosis NOS: 11.1%
Medical Centre GRFD: n=3 (4.2%) (18.0%) All non-affective psychosis:
Utrecht % male: 61.1% COGDIS: n=39 (54.2%) 2 yrs. BLIPS: n=1 (25.0%) 88.9%
UHR+COGDIS: n=32 2 yrs. GRFD: n=1 (33.3%)
Observational study Co-morbidities: 64.6% any (44.4%) Bipolar I disorder with pscchotic
DSM-IV axis-I disorder, 30.9% 2 yrs. UHR+COGDIS: n=7 features: 11.1%
% response: Not depressive disorders (21.9%) All affective psychosis: 11.1%
reported
[92] Mixed: Specialized N=148 SIPS 3.0 and 2 yrs. 2 yrs.: n=28 (18.9%) Conversion: Positive symptom
Velthorst et al., early detection SPI-A (brief) according to PANSS at psychotic
2013 [34] service at the Age: 17.2±3.8 yrs. (11-29 yrs.) Missing observations: intensity 7 days
Adolescent Clinic of APS: n=133 (89.9%) 2 yrs.: n=46 (31.1%)
Dutch the Academic Medical Assigned age group: YOUTH BLIPS: n=10 (7%) Diagnoses: Not reported
Prediction of Centre (AMC) , GRFD: n=6 (4%)
Psychosis University of % male: 64.2% COGDIS: n=55 (37%)
Study (DUPS) Amsterdam and
Department of Child Co-morbidities: Not reported
GE: 2+ and Adolescent
Psychiatry, University
Medical Centre
Utrecht
Observational study
% response: Not
reported
(cumulative)
[93] Mixed, centres with N=103 of TAU condition CAARMS 1.5 yrs. 6 mths.: n=14 (13.4%) Conversion according to CAARMS
van der Gaag and without (2006 version) 1 yr.: n=20 (19.4%)
et al. 2012 [35] specialized early Age: 22.6±5.5 yrs (14-35 yrs.) Missing observations: 1.5 yrs.: n=22 (21.4%) Diagnoses in total sample (n=32
detection services Distribution in total 1.5 yrs.: n=13 conversions):
Dutch Early included. Four sites in Assigned age group: MIX sample (n=201) (17.3%) Schizophrenia: 65.6%
Detection and the Netherlands APS: n=164 (81.6%) Schizoaffective disorder: 3.1%
Intervention recruited participants. In total sample (n=201): % BLIPS: n=3 (1.5%) Brief psychotic disorder: 3.1%
Evaluation male: 48.5% GRFD: n=34 (16.9%) Psychosis NOS: 9.4%
(EDIE-NL) trial Intervention study: All non-affective psychosis:
CBT vs. TAU Co-morbidities of total sample 81.2%
GE: 2+ (n=201): 31.3% anxiety,
% response: 66.5% of 30.8% depressive, and 7.5% Depression with psychotic features:
eligible sample (n=30) personality disorders, 6.5% 12.5%
were randomized ADHD, 6.0% substance-use Bipolar disorder with psychotic
(n=201) disorders, 5.0% PTSD, 3.0% features: 6.3%
oppositional defiant disorder, All affective psychosis: 18.8%
2.5% Asperger
Finland
[101] No specialized early N=52 SIPS 5 yrs. All: n=4 (7.8%) Psychiatric baseline and outcome
Manninen et detection service. diagnoses (ICD-10) were obtained
al. 2013 [36] Reform school Age: 15-18 yrs. APS: n=7 (13.5%) Missing observations: APS: n=1 (14.3%) from the Finnish Hospital Discharge
BLIPS: 0% 0% Register
GE: 2- Observational study Assigned age group: CAD GRFD: 0% None: n=3 (6.7%); 2 Note: Converters might not have
substance-induced sought help after conversion or
% response: 83.9% of % male: 62.3% None: n=45 (86.5%) been treated as outpatients and,
62 eligible residents consequently, might not show in
Co-morbidities: 23.1% with hospital registers
mood disorder, 34.6% with
conduct disorder, 9.6% with Diagnoses: Not reported
substance use, 7.8% with
attention disorder
(cumulative)
[102] No specialized early N=161 SIPS 1 yr. Overall: n=3 (5.7%) Conversion according to SIPS
Lindgren et al. detection service. and/or medical records
2014 [37] Adolescent Age: 16.6±0.9 yrs. (15-18 yrs.) UHR: n=54 (33.5%) Missing observations: APS: n=3 (5.7%)
psychiatric patients in APS: n=53 (98.1%) 8.1% for conversion GRFD: 0% Psychosis NOS: 80%
Helsiniki Helsinki Assigned age group: CAD BLIPS: 0% status All non-affective psychosis: 80%
Prodromal GRFD: n=3 (5.6%) None: n=2 (1.9%)
Study Observational study % male: 33.5% Depression with psychotic features:
None: n=107 (66.5%) 20%
GE: 2+ % response: 75.0% of Co-morbidities: 75.8% mood All affective psychosis: 20%
145 invited patients disorders, 31.7% anxiety
(n=232) completed disorders, 9.3% eating
study protocol disorders, 14.3% substance-
(n=174); n=14 use disorders, 13.7%
excluded for past or disorders usually diagnosed in
present psychosis infancy, childhood, or
adolescence
Switzerland
[99] Specialized early N=64 BSIP Up to 7 yrs. (mean 1 yr.: n=15 (28.3%) Conversion according to BPRS
Riecher- detection service: 5.4 yrs.) 2 yrs.: n=19 (35.8%)
Rössler et al. FEPSY Early Age: 26.5±8.6 yrs. (minimum Of 53 with follow-up: 3 yrs.: n=20 (37.7%) Diagnoses: Not reported
2009 [38] Detection Clinic At the age: 18 yrs.) APS or BLIPS: n=37 Missing observations: > 3 yrs.: n=21 (39.6%)
University Psychiatric (69.8%) 7 yrs.:n=11 (17.2%)
GE: 2+ Outpatient Assigned age group: ADULT Genetic risk: n=2 (3.7%)
Department, Basel APS or BLIPS and
% male: 59.4% genetic risk: n=10
Observational study (18.9%)
Co-morbidities: Not reported Unspecific risk: n=4
% response: 60.4% of (7.6%)
eligible sample
(n=106)
(cumulative)
[100] Specialized early N=148 SIPS 3.0 and SPI-A 1-2 yrs. all CHR: Conversion according to SIPS
Simon et al. detection service: (mean 670 days) 1 yr.: n=7 (12.5%)
2012 [39] Bruderholz Early Age: 14-40 yrs. any CHR: n=99 (66.9%) 2 yrs.: n=10 (23.8%) Schizophrenia: 90%
Psychosis Service UHR: 20.4±5.2 yrs. any UHR: n=73 (49.3%) Missing observations: All non-affective psychosis: 90%
GE: 2+ COPER: 21.7±4.2 yrs. APS: n=68 (46.0%) UHR UHR, 1 yr.: n=7 (9.6%)
Observational study none: 21.6±5.0 yrs. BLIPS: n=3 (2.0%) 1 yr.: n=17 (23.3%) UHR, 2 yrs.: n=10 (13.7%) All affective psychosis: 10%
GRFD: n=2 (1.4%) 2 yrs.: n=40 (54.8%) APS, 1 yr.: n= 7 (10.3%)
% response: Not Assigned age group: MIX COPER (UHR criteria APS, 2 yrs.: n=10 (14.7%)
reported excluded): n=26 (17.6%) COPER
% male: 67.6% 1 yr.: n=11 (42.3%) all other CHR criteria: 0%
none: n=49 (33.1%) 2 yrs.: n=10 (38.5%)
Co-morbidities in CHR
sample: 65.7% mood/anxiety none
disorders, 8.1% adjustment 1 yr.: n=23 (46.9%)
disorder, 6.1% dissociative 2 yrs.: n=31 (63.3%)
disorder
Multiple or other European countries
[103] Mixed, centres with N=245 SPI-A (brief) and 1.5 yrs. 1.5 yrs.: n=37 (15.1%) Conversion assessed with SIPS:
Ruhrmann et and without SIPS (3.0 with modified 1.5 yrs. only COGDIS: 5% any
al. 2010 [40]; specialized early Age: 23.0±5.2yrs. (16-35 yrs.) GRFD) Missing observations: 1.5 yrs. only UHR: 18% positive item = 6 for >7 days
Salokangas et detection services 1.5 yrs.: n=62 1.5 yrs. COGDIS+UHR:
al. 2012 [41] included. Assigned age group: ADULT Only COGDIS: 25 (25.3%) 22% Schizophrenia: 62.2%
(10.2%) Schizophreniform disorder: 8.1%
European Observational study % male: 55.9% Only UHR: 74 (30.2%) Brief psychotic disorder: 5.4%
Prediction of COGDIS+UHR: 146 Schizoaffective disorder: 8.1%
Psychosis % response: 48% of Co-morbidities: 62.0% any and (59.6%) All non-affective psychosis:
Study (EPOS) eligible sample 22.4% even 2-3 current axis-I 83.8%
incl. Germany, (n=513) disorders; 39.2% anxiety
Finland, UK disorder, 34.3% unipolar Mood disorder with psychotic
and the depressive disorders, 4.1% features: 16.2%
Netherlands bipolar disorders, 6.5% All affective psychosis: 16.2%
somatoform disorders, 3.2%
GE: 2+ other disorders
(cumulative)
[104] No specialized early N=40 of placebo condition PANSS 1 yr. 1 yr.: n=11 (27.5%) Conversion according to BPRS and
Amminger et detection service: PANSS
al. 2010 [42] First episode psycho- Age: 16.0±1.7 yrs. (13-25 yrs.) APS: n=22 (55.0%) Missing observations:
sis detection unit of BLIPS: n=3 (7.5%) 1 yr.: n=2 (5.0%) Schizophrenia: 72.7%
GE: 2+ the Department of Assigned age group: CAD GRFD: 0% Schizophreniform disorder: 9.1%
Child and Adolescent APS+BLIPS: n=13 Schizoaffective disorder: 9.1%
Psychiatry, Vienna, % male: 32.5% (32.5%) All non-affective psychosis:
Austria APS+GRFD: n=2 (5.0%) 90.9%
Intervention study: Bipolar I disorder with psychotic
Omega-3 fatty acids features: 9.1%
vs. placebo All affective psychosis: 9.1%
% response: 76.4% of
eligible sample
(n=106) were
randomized (n=81)
[105] Specialized early N=40 CAARMS (Italian 1 yr. 1 yr.: n=9 (22.5%) Conversion according to CAARMS
Fusar-Poli et detection service: translation)
al. 2012 [43] Programma 2000, Age: 20.7±5.3 yrs. (15-35 yrs.) Missing observations: Diagnoses: Not reported
Milan, Italy APS: n=36 (90%) 0%
GE: 2+ Assigned age group: MIX BLIPS: n=1 (2.5%)
Observational study GRFD: n=3 (7.5%)
% male: 47.5%
% response: Not
reported Co-morbidities: Not reported
(cumulative)
[106] Specialized early N=61 SIPS 3.0 3 yrs. 1 yr.: n=11 (18.0%) Conversion according to SIPS
Lemos- detection service: 3 yrs.: n=14 (23.0%)
Giráldez et al. Prevention program Age: 21.7±3.8 yrs. (15-31 yrs.) APS: n=52 (85.2%) Missing observations: Schizophrenia: 78.6%
2009 [44] for psychosis (P3), BLIPS: n=3 (4.9%) 3 yrs.: n=16 (26.2%) Schizophreniform disorder: 7.1%
Hospital Sierrallana, Assigned age group: ADULT GRFD: n=6 (9.8%) Substance-induced psychosis:
GE: 2- Torrelavega, Spain 14.3%
% male: 65.6% All non-affective psychosis:
Observational study 100%
Various recruitment Converters significantly older at
sources incl. website, baseline.
no participant was
help-seeking
% response: Not
applicable, only
contact with persons
willing to participate
[107] No specialized early N=97 CAARMS (before 2006 1 yr. 1 yr.: n=31 (32.0%) Conversion according to CAARMS
Kiss et al. detection service. version)
2012 [45]; Outpatient units of the Age: Not reported Missing observations: Psychotic disorders (schizophrenia,
Letter to the University of Szeged, UHR: n=97 (100%) Not reported schizophreniform disorder,
Editor Bács-Kiskun Country No age group assignment. psychotic mood disorder): 100%
Hospital, Kecskemét,
GE: 2+ & National Psychiatry Gender: Not reported
Center, Semmelweis
University, Budapest, Co-morbidities: Not reported
Hungary
Observational study
% response: Not
reported
(cumulative)
[108] Specialized early N=81 CAARMS 2006 version 3 yrs. Overall: n=15 (18.5%) Conversion according to CAARMS
Kotlicka- detection service:
Antczak et al. Programme of Age: 15-29 yrs. UHR: n=81 (100%) (all Missing observations: Diagnoses: Not reported
2014 [46]; Recognition and plus functional decline 3 yrs.: n=16 (19.7%)
Brief report Therapy (PORT), Assigned age group: MIX within past 12 mths.)
Central Clinical
GE: 2+ Hospital of Lodz, Gender: Not reported
Poland
Observational study
% response: Not
reported
Asian countries
[109] Specialized early N=62 CAARMS (before 2006 6 mths. 3 mths.: n=16 (25.8%) Conversion according to CAARMS
Lam et al. detection service: version) and PANSS 6 mths.: n=18 (29.0%)
2006 [47] Early Assessment Age: 16.2±3.7 yrs. (6.9-23.5 Missing observations: Schizophreniform disorder: 61.1%
Service for Young yrs.) APS: n=51 (82.2%) n=9 (14.5%) Schizoaffective disorder: 11.1%
GE: 2+ People with psychosis BLIPS: n=12 (19.4%) Psychosis NOS: 16.6%
(EASY), Hong-Kong, Assigned age group: YOUTH GRFD: n=12 (19.4%) All non-affective psychosis:
China 88.8%
% male: 58.1%
Observational study Bipolar disorder with psychotic
Co-morbidities: 93.5% had a features: 5.6%
% response: 92.5% of non-psychotic DSM-IV Depression with psychotic features:
eligible sample (n=67) diagnosis, 6.5% were 5.6%
diagnosed with brief psychosis All affective psychosis: 11.2%
or substance-use psychosis
due to BLIPS
(cumulative)
[113] No specialized early N=89 SIPS 3.0 2 yrs. 2 yrs.: n=14 (15.7%) Conversion according to SIPS
Zhang et al. detection service:
2014 [48] Shanghai Age: 25.89±7.54 yrs. (15-45 yrs.) APS: n=65 (73%) Missing observations: Schizophrenia: 85.7%
Psychotherapy and BLIPS: n=3 (3.4%) 2 yrs.: n=36 (40.5%) Other psychotic disorder according
GE: 2+ Psychological Assigned age group: ADULT GRFD: n=25 (28.1%) to POPS: 14.3%
Counselling Centre All non-affective psychosis:
(SPCC), China % male: 50.6% 85.7%-100%
Observational study Co-morbidities: 24.7% mood or

anxiety disorder, 3.4% stress
% response: 82.3% of related disorder, 7.9% other
invited screening disorder or not yet determined
positives on the PQ-B
(n=1681) were
interviewed (n=1384)
[110] No specialized early N=667 CAARMS (before 2006 6 mths. 6 mths., UHR: n=6 (3.5%) Conversion according to PANSS
Lee et al. 2013 detection service: version) and SCID
[49] Longitudinal Youth At- Age: UHR: 21.3±3.5 yrs.; Missing observations: 6 mths., none: 0%
Risk Study (LYRIKS), none: 21.7±3.4 yrs. (14-29 yrs.) UHR: n=173 (25.9%) of none; 2 yrs.: n=148 Diagnoses: Not reported
GE: 2+ Singapore these 76.3% help- (30.0%)
Assigned age group: ADULT seekers
Observational study APS: n=144 (83.7%) UHR; 2 yrs.: 0%
of both help-seeking % male: 60.1% BLIPS: n=6 (3.5%)
and non-help-seeking GRFD: n=49 (28.5%)
persons Co-morbidities:
in UHR: 78.0% any axis-I; 65.3% none: n=494 (74.1%) of
% response: 39.1% of mood, and 22.5% anxiety disorders, these 14.2% help-
invited persons 11.0% OCD, 19.6% substance-use seekers
(n=2368) agreed to disorders, 5.8% adjustment
assessment (n=926); disorders incl. PTSD
n=667 were accepted
into the study in none: 18.8% any axis-I disorder;
12.2% mood, and 3.4% anxiety
disorders, 1.8% OCD, 5.1%
substance-use disorders, 1.2%
adjustment disorders incl. PTSD
(cumulative)
[111] Specialized early N=106 CAARMS (before 2006 7.4 yrs. Overall: n=14 (13.2%) Conversion according to CAARMS
Katsura et al. detection service: version) (mean=3.2 yrs.;
2014 [50] Sendai ARMS and Age: 20.0±4.3 yrs. Mdn=2.7 yrs.; min. 1 1 yr.: n=10 (9.4%) Schizophrenia: 57.1%
first-episode (SAFE) (14-35 yrs.) APS: n=99 (93.4%) yr.) 2 yrs.: n=13 (12.3%) Schizophreniform disorder: 7.1%
GE: 2+ clinic, Sendai, Japan BLIPS: n=4 (3.8%) 3 yrs.: n=14 (13.2%) Delusional disorder: 7.1%
Assigned age group: MIX GRFD: n=3 (2.8%) Missing observations: Psychosis NOS: 28.6%
Observational study 1 yr.: n=23 (21.7%) APS: n=12 (18.3%) All non-affective psychosis:
% male: 37.7% 1 yr.: Not reported BLIPS: n=2 (50.0%) 100%
% response: 95.5% of GRFD: 0%
eligible sample Co-morbidities: Not reported
(n=111)
[112] No specialized early N=37 SIPS 2 yrs. Overall: n=6 (16.2%) Conversion according to SIPS
Koike et al. detection service:
2013 [51] Outpatient and Age: 21.3±3.6 yrs. (15-30 yrs.) APS: n=32 (86.5%) Missing observations: 6 mths.: n=2 (5.4%) Diagnoses: Not reported
inpatient units of the BLIPS: n=3 (8.1%) 6 mths.: n=10 12 mths.: n=2 (5.4%)
GE: 2+ University of Tokyo Assigned age group: ADULT GRFD: n=10 (27%) (27.0%) 24 mths.: n=6 (16.2%)
Hospital, Japan 12 mths.: n=13
% male: 54.1% (37.1%)
Observational study 24 mths.: n=20
Co-morbidities: Not reported (54.1%)
% response: 74% of
eligible sample (n=50)
[114] Specialized early N=78 CAARMS (before 2006 7 yrs. Overall: n=14 (20.9%) Conversion according to DSM-IV
Kim et al. 2012 detection service: version) (mean time to
[52] Seoul Youth Clinic, Age: 21.3±4.2 yrs. conversion: 412 days; APS: n=13 (18.3%) Schizophrenia: n=10
South Korea APS: n=71 (91%) 32-1127 days) BLIPS: n=1 (100%) All non-affective psychosis:
GE: 2+ Assigned age group: MIX BLIPS: n=1 (1.3%) GRFD: n=1 (7.7%) 71.4%
Observational study GRFD: n=13 (16.7%) Missing observations:
% male: 87.2% Overall: n=11 (14.1%) Bipolar I with psychotic features:
% response: Not n=4
reported Co-morbidities: 61% mood All affective psychosis: 28.6%
disorders, 17% anxiety disorders,
3% other disorders
STable 2 Effect sizes (Ei, 𝐸 or 𝐸 ∗ ) at different follow-ups tx in samples meeting COPER,

single UHR criteria (each irrespective of the potential presence of other CHR
criteria) or certain UHR-COGDIS combinations as well as in CHR-negative
samples
Study tx N Ei, 𝑬 or 𝑬∗ 95% CIs Q (df) I2 z
Cognitive-perceptive basic symptoms (COPER)
[25] 6,a 6 mth. 146 0.158 0.129; 0.187
[25] 6,a 1 yr. 146 0.247 0.212; 0.282
5,a
[20] 1 yr. 106 0.198 0.160; 0.236
4,a
[23] 1 yr. 64 0.139 0.097; 0.181
6,b #
[39] 1 yr. 26 0 0; 0
pooled 1 yr. 0.144 0.072; 0.215 81.169 (3) 96.3% 1.976*
[25] 6,a 2 yrs. 146 0.329 0.291; 0.367
[20] 5,a 2 yrs. 106 0.368 0.322; 0.414
[23] 4,a 2 yrs. 64 0.154 0.110; 0.198
[39] 6,b 2 yrs. 26 0# 0; 0
pooled 2 yrs. 0.211 0.107; 0.315 142.066 (3) 97.9% 1.989*
5,a
[20] 3 yrs. 106 0.500 0.452; 0.548
5,a
[20] 4 yrs. 106 0.557 0.510; 0.604
[20] 5,a 4 yrs. 106 0.651 0.606; 0.696
Note: Attenuated and transient psychotic symptoms, and UHR criteria, respectively, were
excluded in [23] and [39].
Upper number indicates scale used for the assessment of CHR criteria (4: ERIraos; 5:
BSABS; and 6: SPI-A)
Upper small letter indicates age group of sample (a: ADULT; b: MIX; c: YOUTH; and d: CAD)
F: according to fixed-effects model
* z  1.96: significant on 5% level; ** z  2.58: significant on 1% level
#
Because inclusion of extreme Ei values, i.e., 0 or 1, would cause an undue division by 0 in
the calculation of 𝐸 or 𝐸 ∗ , 0 was replaced by 0.001 and 1 by 0.999 in their calculation.
STable 2 cont. (1)
Study tx N Ei, 𝑬 or 𝑬∗ 95% CIs Q (df) I2 Z

Respective examined clinical high risk criteria not fulfilled (CHR-negative)
[22] 1+6,a;UHR/COGDIS 6 mth. 52 0.019 -0.001; 0.039
[49] 2,a;UHR 6 mth. 494 0# 0; 0
[7,8] 2,c;UHR 6 mth. 173 0.006 0.000; 0.011
pooled F 6 mth. 0.001 -0.0002; 0.0002 2.020 (2) 1.0% 0.148
[22] 1+6,a;UHR/COGDIS 1 yr. 52 0.038 0.012; 0.064
[39] 1,b;UHR/COPER 1 yr. 49 0# 0; 0
[37] 1,d;UHR 1 yr. 107 0.019 0.005; 0.033
F
pooled 1 yr. 0.001 -0.001; 0.003 3.650 (2) 45.2% 0.394
1+6,a;UHR/COGDIS
[22] 2 yrs. 52 0.077 0.040; 0.114
1,b;UHR/COPER #
[39] 2 yrs. 49 0 0; 0
[7,8] 2,c;UHR 2 yrs. 173 0.012 0.004; 0.020
pooled 2 yrs. 0.009 0.001; 0.017 6.297 (2) 68.2% 0.994
[22] 1+6,a;UHR/COGDIS 3 yrs. 52 0.115 0.071; 0.159
[22] 1+6,a;UHR/COGDIS 4 yrs. 52 0.115 0.071; 0.159
[22] 1+6,a;UHR/COGDIS 4 yrs. 52 0.154 0.105; 0.203
[36] 1,d;UHR 4 yrs. 45 0.067 0.030; 0.104
pooled F 4 yrs. 0.098 0.069; 0.128 1.941 (1) 48.5% 3.279**
Upper number indicates scale used for the assessment of CHR criteria (1: SIPS; 2:
CAARMS; 3: CAARMS 2006 version; 4: other scale; 5: BSABS; and 6: SPI-A)
#
STable 2 cont. (2)

Attenuated psychotic symptoms (APS) criterion
[5] 2,b 6 mth. 664 0.076 0.067; 0.085
[17] 1,a 6 mth. 24 0.125 0.059; 0.191
pooled F 6 mth. 0.077 0.068; 0.086 0.512 (1) 0% 8.189**
[39] 1,b 1 yr. 68 0.103 0.066; 0.140
[37] 1,d 1 yr. 53 0.057 0.026; 0.088
F
pooled 1 yr. 0.076 0.052; 0.100 0.882 (1) 0% 3.154**
1,b
[39] 2 yrs. 68 0.147 0.103; 0.191
1,d
[33] 2 yrs. 65 0.138 0.096; 0.180
[52] 1,c 2 yrs. 71 0.183 0.138; 0.228
[14] 1,c 2 yrs. 65 0.262 0.208; 0.315
pooled F 2 yrs. 0.174 0.151; 0.197 3.704 (3) 19.0% 7.517**
[10] 2,b 3 yrs. 43 0.006 0.437; 0.587
1,b
[50] 3 yrs. 99 0.121 0.088; 0.154
1,c
[12] 3 yrs. 282 0.280 0.249; 0.311
pooled 3 yrs. 0.291 0.205; 0.378 27.012 (2) 92.6% 3.296 **
[15] 1,d 4 yrs. 101 0.149 0.115; 0.183
[36] 1,d 4 yrs. 7 0.143 0.013; 0.273
pooled F 4 yrs. 0.149 0.116; 0.181 0.002 (1) 0% 4.435**
#
STable 2 cont.(3)

Transient psychotic symptoms (BLIPS) criterion
[5] 2,c 6 mth. 36 0.133 0.077; 0.189
[39] 1,b 1 yr. 3 0# 0; 0
[52] 2,b 2 yrs. 1 1# 0.606; 1.194
[14] 1,c 2 yrs. 17 0.588 0.472; 0.704
[33] 1,d 2 yrs. 4 0.250 0.038; 0.462
1,b #
[39] 2 yrs. 3 0 0; 0
pooled 2 yrs. 0.466 0.188; 0.744 31.740 (3) 90.5% 1.642
2,c
[10] 3 yrs. 23 0.609 0.509; 0.709
[50] 2,b 3 yrs. 4 0.500 0.255; 0.745
[12] 1,c 3 yrs. 7 0.429 0.246; 0.612
pooled F 3 yrs. 0.518 0.379; 0.656 5.421(2) 63.1% 3.659**
Genetic risk and functional decline (GRFD) criterion
2,c
[5] 6 mth. 209 0.062 0.045; 0.079
1,d #
[37] 1 yr. 3 0 0; 0
1,b #
[39] 1 yr. 2 0 0; 0
pooled F 1 yr. 0 0; 0 0 (1) 0% Div/0
[14] 1,c 2 yrs. 2 0# 0; 0
[39] 1,b 2 yrs. 2 0# 0; 0
[33] 1,d 2 yrs. 3 0.333 -0.511; 1.177
[52] 2,b 2 yrs. 13 0.077 0.004; 0.150
F
pooled 2 yrs. 0.019 -0.012; 0.050 0.760 (3) 0% 0.588
2,c
[10] 3 yrs. 19 0.105 0.036; 0.174
1,c #
[12] 3 yrs. 2 0 0; 0
[50] 2,b 3 yrs. 3 0# 0; 0
pooled F 3 yrs. 0.014 -0.008; 0.037 1.802 (2) 0% 0.637
#
STable 2 cont.(4)

Ultra-high risk (UHR) criteria and/or Cognitive disturbances (COGDIS)
[22] 1+6,a 6 mth. 194 0.149 0.123; 0.175
[22] 1+6,a 1 yr. 194 0.258 0.227; 0.289
[33] 1+6,d 1 yr. 72 0.097 0.063; 0.131
pooled 1 yr. 0.178 0.099; 0.257 11.782 (1) 91.5% 2.213*
[22] 1+6,a 2 yrs. 194 0.335 0.301; 0.369
1+6,a
[40] 2 yrs. 245 0.151 0.129; 0.173
1+6,a
[33] 2 yrs. 72 0.125 0.087; 0.163
1+6,a
[34] 2 yrs. 148 0.189 0.158; 0.220
pooled 2 yrs. 0.199 0.157; 0.242 23.380 (3) 87.2% 4.596**
[22] 1+6,a 3 yrs. 194 0.371 0.337; 0.405
[22] 1+6,a 4 yrs. 194 0.387 0.353; 0.421
[24] 2+5,a 4 yrs. 45 0.311 0.243; 0.379
F
pooled 4 yrs. 0.372 0.341; 0.402 0.963 (1) 0% 12.000**
4 yrs.
1+6,a
[22] 194 0.402 0.368; 0.436
Ultra-high risk (UHR) criteria plus Cognitive disturbances (COGDIS)
[22] 1+6,a 6 mth. 127 0.165 0.132; 0.198
[22] 1+6,a 1 yr. 127 0.299 0.259; 0.339
[22] 1+6,a 2 yrs. 127 0.409 0.366; 0.452
[40] 1+6,a 2 yrs. 146 0.171 0.140; 0.202
[33] 1+6,d 2 yrs. 32 0.219 0.148; 0.290
pooled 2 yrs. 0.267 0.185; 0.350 19.727 (2) 89.9% 3.174**
1+6,a
[22] 3 yrs. 127 0.449 0.406; 0.492
1+6,a
[22] 4 yrs. 127 0.472 0.428; 0.516
[22] 1+6,a 4 yrs. 127 0.496 0.452; 0.540
#
STable 3 Pairwise comparison of conversion rates at different follow-ups tx in samples

meeting single UHR criteria (each irrespective of the potential presence of other
CHR criteria) and in CHR-negative samples (one-dimensional 2 tests with df = 1)
Study BLIPS GRFD CHR-negative
APS 6 mths.: 2 = 1.493 6 mths.: 2 = 0.162 6 mths.: 2 = 7.694 **
1 yr.: 2 = 7.700 ** 1 yr.: 2 = 7.700 ** 1 yr.: 2 = 7.493 **
2 yrs.: 2 = 13.323 *** 2 yrs.: 2 = 12.675 *** 2 yrs.: 2 = 17.102 ***
3 yrs.: 2 = 9.459 ** 3 yrs.: 2 = 20.486 *** 3 yrs.: 2 = 4.933 *
4 yrs.: no data 4 yrs.: no data 4 yrs.: no data
4 yrs.: no data 4 yrs.: no data 4 yrs.: 2 = 1.016 §
BLIPS 6 mths.: 2 = 2.585 6 mths.: 2 = 13.294 ***
1 yr.: 2 = 0 1 yr.: 2 = 0.080
2 yrs.: 2 = 41.678 *** 2 yrs.: 2 = 44.083 ***
3 yrs.: 2 = 47.092 *** 3 yrs.: 2 = 25.570 ***
4 yrs.: no data 4 yrs.: no data
4 yrs.: no data 4 yrs.: no data
GRFD 6 mths.: 2 = 6.194 *
1 yr.: 2 = 0.080
2 yrs.: 2 = 1.521
3 yrs.: 2 = 7.482 **
4 yrs.: no data
4 yrs.: no data
APS: attenuated psychotic symptoms criterion; BLIPS: transient psychotic symptoms
criterion; GRFD: genetic risk and functional decline criterion
* p  0.05; ** p  0.01; *** p  0.001
§
both comparative APS samples are CAD samples
STable 4 Pairwise comparison of conversion rates at different follow-ups tx in UHR

samples assessed with different scales and, relatedly, according to different UHR
criteria, and total sample (one-dimensional 2 tests with df = 1)
Study SIPS CAARMS early CAARMS 2006 version
versions
Total UHR 6 mths.:  = 0.2332
6 mths.: 2 = 0.027 6 mths.: 2 = 0.080
sample 1 yr.: 2 = 0.560 1 yr.: 2 < 0.001 1 yr.: 2 = 0.551
2 yrs.: 2 = 0.002 2 yrs.: 2 = 0.370 2 yrs.: 2 = 0.221
3 yrs.: 2 = 0.054 3 yrs.: 2 = 0.017 3 yrs.: 2 = 2.361
4 yrs.: 2 = 0.013 4 yrs.: 2 = 0.968 4 yrs.: 2 = 2.519
4 yrs.: 2 = 0.379 4 yrs.: 2 = 0.001 4 yrs.: no data
CAARMS 6 mths.: 2 = 0.120 6 mths.: 2 = 0.015
early
versions 1 yr.: 2 = 0.578 1 yr.: 2 = 0.570
2 yrs.: 2 = 0.314 2 yrs.: 2 = 1.309
3 yrs.: 2 = 0.011 3 yrs.: 2 = 2.769 °
4 yrs.: 2 = 1.202 4 yrs.: 2 = 6.453 *
4 yrs.: 2 = 0.351 4 yrs.: no data
CAARMS 6 mths.: 2 = 0.051
2006
version 1 yr.: 2 < 0.001
2 yrs.: 2 = 0.345
3 yrs.: 2 = 3.113 °
4 yrs.: 2 = 2.178
4 yrs.: no data
SIPS: Structured Interview for Psychosis-Risk Syndromes [53]
CAARMS early versions: Comprehensive Assessment for At-Risk Mental States, versions
before 2006 versions [54]
CAARMS 2006 version: Comprehensive Assessment for At-Risk Mental States, 2006 version
[55]
° p  0.10; * p  0.05
STable 5 Pairwise comparison of conversion rates at different follow-ups tx in UHR

samples of different age groups and total sample (one-dimensional 2 tests with
df = 1)
Study CAD YOUTH ADULT
Total UHR 6 mths.: 2 = 3.003 ° 6 mths.: 2 = 0.760 6 mths.: 2 = 0.347
sample 1 yr.: 2 = 1.235 1 yr.: 2 = 0.970 1 yr.: 2 = 0.273
2 yrs.: 2 = 2.718 ° 2 yrs.: 2 = 0.129 2 yrs.: 2 = 0.549
3 yrs.: no data 3 yrs.: 2 = 0.509 3 yrs.: 2 = 0.138
4 yrs.: no data 4 yrs.: 2 = 0.968 4 yrs.: 2 < 0.001
4 yrs.: 2 = 10.045 ** 4 yrs.: 2 = 0.001 4 yrs.: 2 = 0.158
YOUTH 6 mths.: 2 = 6.447 * 6 mths.: 2 = 2.148
1 yr.: 2 = 4.275 * 1 yr.: 2 = 0.216
2 yrs.: 2 = 3.978 * 2 yrs.: 2 = 0.147
3 yrs.: no data 3 yrs.: 2 = 0.117
4 yrs.: no data 4 yrs.: 2 = 0.914
4 yrs.: 2 = 9.838 ** 4 yrs.: 2 = 0.187
ADULT 6 mths.: 2 = 1.388
1 yr.: 2 = 2.627
2 yrs.: 2 = 5.568 *
3 yrs.: no data
4 yrs.: no data
4 yrs.: 2 = 12.526 ***
CAD: almost entirely minors (18 years); YOUTH: 50% minors; ADULT: almost entirely
adults
° p  0.10; * p  0.05; ** p<0.01; *** p<0.001

S.8. References of Supplementary Material
[1] Eisend M. Metaanalyse. Meiring: Rainer Hampp Verlag. 2014.

[2] Higgins J, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-
analyses. BMJ. 2003;327(7414):557-60.
[3] Koricheva J, Gurevitch J, Mengersen K (eds.). Handbook of meta-analysis in ecology
and evolution. Princeton, NJ: Princeton University Press. 2013.
[4] Nelson B, Yuen HP, Wood SJ, Lin A, Spiliotacopoulos D, Bruxner A, Broussard C,
Simmons M, Foley DL, Brewer WJ, Francey SM, Amminger GP, Thompson A, McGorry
PD, Yung AR. Long-term follow-up of a group at ultra high risk ("prodromal") for
psychosis: the PACE 400 study. JAMA Psychiatry. 2013;70(8):793-802.
[5] Nelson B, Yuen K, Yung AR. Ultra high risk (UHR) for psychosis criteria: are there
different levels of risk for transition to psychosis? Schizophr Res. 2011;125(1):62-
[6] Nelson B, Thompson A, Yung AR. Basic self-disturbance predicts psychosis onset in the
ultra high risk for psychosis "prodromal" population. Schizophr Bull. 2012;38(6):1277-87.
[7] Yung AR, Stanford C, Cosgrave E, Killackey E, Phillips L, Nelson B, McGorry PD.
Testing the ultra high risk (prodromal) criteria for the prediction of psychosis in a clinical
sample of young people. Schizophr Res. 2006;84(1):57-66.
[8] Yung AR, Nelson B, Stanford C, Simmons MB, Cosgrave EM, Killackey E, Phillips LJ,
Bechdolf A, Buckby J, McGorry PD. Validation of "prodromal" criteria to detect
individuals at ultra high risk of psychosis: 2 year follow-up. Schizophr Res. 2008;105(1-
3):10-7.
[9] Bechdolf A, Thompson A, Nelson B, Cotton S, Simmons MB, Amminger GP, Leicester S,
Francey SM, McNab C, Krstev H, Sidis A, McGorry PD, Yung AR. Experience of trauma
and conversion to psychosis in an ultra-high-risk (prodromal) group. Acta Psychiatr
Scand. 2010;121(5):377-84.
[10] Mason O, Startup M, Halpin S, Schall U, Conrad A, Carr V. Risk factors for transition to
first episode psychosis among individuals with 'at-risk mental states'. Schizophr Res.
2004;71(2-3):227-37.
[11] Welsh P, Tiffin PA. The 'at-risk mental state' for psychosis in adolescents: clinical
presentation, transition and remission. Child Psychiatry Hum Dev. 2014;45(1):90
[12] Cannon TD, Cadenhead K, Cornblatt B, Woods SW, Addington J, Walker E, Seidman
LJ, Perkins D, Tsuang M, McGlashan T, Heinssen R. Prediction of psychosis in youth at
high clinical risk: a multisite longitudinal study in North America. Arch Gen Psychiatry.
2008;65(1):28-37.
[13] Addington J, Cadenhead KS, Cannon TD, Cornblatt B, McGlashan TH, Perkins DO,
Seidman LJ, Tsuang M, Walker EF, Woods SW, Heinssen R; North American Prodrome
Longitudinal Study. North American Prodrome Longitudinal Study: a collaborative
multisite approach to prodromal schizophrenia research. Schizophr Bull. 2007;33(3):665-

72.
[14] Schlosser DA, Jacobson S, Chen Q, Sugar CA, Niendam TA, Li G, Bearden CE, Cannon
TD. Recovery from an at-risk state: clinical and functional outcomes of putatively
prodromal youth who do not develop psychosis. Schizophr Bull. 2012;38(6):1225-33.
[15] Carrión RE, McLaughlin D, Goldberg TE, Auther AM, Olsen RH, Olvet DM, Correll CU,
Cornblatt BA. Prediction of functional outcome in individuals at clinical high risk for
psychosis. JAMA Psychiatry. 2013;70(11):1133-42.
[16] Woodberry KA, Seidman LJ, Giuliano AJ, Verdi MB, Cook WL, McFarlane WR.
Neuropsychological profiles in individuals at clinical high risk for psychosis: relationship
to psychosis and intelligence. Schizophr Res. 2010;123(2-3):188-98.
[17] Addington J, Epstein I, Liu L, French P, Boydell KM, Zipursky RB. A randomized
controlled trial of cognitive behavioral therapy for individuals at clinical high risk of
psychosis. Schizophr Res. 2011;125(1):54-61.
[18] Buchy L, Perkins D, Woods SW, Liu L, Addington J. Impact of substance use on
conversion to psychosis in youth at clinical high risk of psychosis. Schizophr Res.
2014;156(2-3):277-80.
[19] Kayser J, Tenke CE, Kroppmann CJ, Alschuler DM, Ben-David S, Fekri S, Bruder GE,
Corcoran CM. Olfaction in the psychosis prodrome: electrophysiological and behavioral
measures of odor detection. Int J Psychophysiol. 2013;90(2):190-206.
[20] Klosterkötter J, Hellmich M, Steinmeyer EM, Schultze-Lutter F. Diagnosing
schizophrenia in the initial prodromal phase. Arch Gen Psychiatry. 2001;58(2):158-64.
[21] Schultze-Lutter F, Ruhrmann S, Klosterkötter J. In: Johannessen JO, Martindale B,
Cullberg J, editors.? Evolving Psychosis. Different Stages, Different Treatments. London,
New York: Routledge. 2006;104-123.
[22] Schultze-Lutter F, Klosterkötter J, Ruhrmann S. Improving the clinical prediction of
psychosis by combining ultra-high risk criteria and cognitive basic symptoms. Schizophr
Res. 2014;154(1-3):100-6.
[23] Bechdolf A, Wagner M, Ruhrmann S, Harrigan S, Putzfeld V, Pukrop R, Brockhaus-
Dumke A, Berning J, Janssen B, Decker P, Bottlender R, Maurer K, Möller HJ, Gaebel
W, Häfner H, Maier W, Klosterkötter J. Preventing progression to first-episode psychosis
in early initial prodromal states. Br J Psychiatry. 2012;200(1):22-9.
[24] Koutsouleris N, Meisenzahl EM, Davatzikos C, Bottlender R, Frodl T, Scheuerecker J,
Schmitt G, Zetzsche T, Decker P, Reiser M, Möller HJ, Gaser C. Use of neuroanatomical
pattern classification to identify subjects in at-risk mental states of psychosis and predict
disease transition. Arch Gen Psychiatry. 2009;66(7):700-12.
[25] Schultze-Lutter F, Klosterkötter J, Picker H, Steinmeyer EM, Ruhrmann S. Predicting

first-episode psychosis by basic symptom criteria. Clin Neuropsychiatry. 2007;4(1):11-
22.
[26] Bodatsch M, Ruhrmann S, Wagner M, Müller R, Schultze-Lutter F, Frommann I,
Brinkmeyer J, Gaebel W, Maier W, Klosterkötter J, Brockhaus-Dumke A. Prediction of
psychosis by mismatch negativity. Biol Psychiatry. 2011;69(10):959-66.
[27] Ruhrmann S, Bechdolf A, Kühn KU, Wagner M, Schultze-Lutter F, Janssen B, Maurer K,
Häfner H, Gaebel W, Möller HJ, Maier W, Klosterkötter J; LIPS study group. Acute
effects of treatment for prodromal symptoms for people putatively in a late initial
prodromal state of psychosis. Br J Psychiatry. 2007;51(Suppl):88-95.
[28] Fusar-Poli P, Byrne M, Badger S, Valmaggia LR, McGuire PK. Outreach and support in
south London (OASIS), 2001-2011: ten years of early diagnosis and treatment for young
individuals at high clinical risk for psychosis. Eur Psychiatry. 2013;28(5):315-26.
[29] Morrison AP, French P, Walford L, Lewis SW, Kilcommons A, Green J, Parker S, Bentall
RP. Cognitive therapy for the prevention of psychosis in people at ultra-high risk:
randomised controlled trial. Br J Psychiatry. 2004;185:291-7.
[30] Morrison AP, French P, Parker S, Roberts M, Stevens H, Bentall RP, Lewis SW. Three-
year follow-up of a randomized controlled trial of cognitive therapy for the prevention of
psychosis in people at ultrahigh risk. Schizophr Bull. 2007;33(3):682-7.
[31] Morrison AP, Stewart SL, French P, Bentall RP, Birchwood M, Byrne R, Davies LM,
Fowler D, Gumley AI, Jones PB, Lewis SW, Murray GK, Patterson P, Dunn G. Early
detection and intervention evaluation for people at high-risk of psychosis-2 (EDIE-2): trial
rationale, design and baseline characteristics. Early Interv Psychiatry. 2011;5(1):24-32.
[32] Morrison AP, French P, Stewart SL, Birchwood M, Fowler D, Gumley AI, Jones PB,
Bentall RP, Lewis SW, Murray GK, Patterson P, Brunet K, Conroy J, Parker S, Reilly T,
Byrne R, Davies LM, Dunn G. Early detection and intervention evaluation for people at
risk of psychosis: multisite randomised controlled trial. BMJ. 2012;344:e2233.
[33] Ziermans TB, Schothorst PF, Sprong M, van Engeland H. Transition and remission in
adolescents at ultra-high risk for psychosis. Schizophr Res. 2011;126(1-3):58-64.
[34] Velthorst E, Derks EM, Schothorst P, Becker H, Durston S, Ziermans T, Nieman DH, de
Haan L. Quantitative and qualitative symptomatic differences in individuals at Ultra-High
Risk for psychosis and healthy controls. Psychiatry Res. 2013;210(2):432-7.
[35] van der Gaag M, Nieman DH, Rietdijk J, Dragt S, Ising HK, Klaassen RM, Koeter M,
Cuijpers P, Wunderink L, Linszen DH. Cognitive behavioral therapy for subjects at
ultrahigh risk for developing psychosis: a randomized controlled clinical trial. Schizophr
Bull. 2012;38(6):1180-8.
[36] Manninen M, Lindgren M, Therman S, Huttunen M, Ebeling H, Moilanen I, Suvisaari J.

Clinical high-risk state does not predict later psychosis in a delinquent adolescent
population. Early Interv Psychiatry. 2014;8(1):87-90.
[37] Lindgren M, Manninen M, Kalska H, Mustonen U, Laajasalo T, Moilanen K, Huttunen M,
Cannon TD, Suvisaari J, Therman S. Predicting psychosis in a general adolescent
psychiatric sample. Schizophr Res doi: 10.1016/j.schres.2014.06.028.
[38] Riecher-Rössler A, Pflueger MO, Aston J, Borgwardt SJ, Brewer WJ, Gschwandtner U,
Stieglitz RD. Efficacy of using cognitive status in predicting psychosis: a 7-year follow-
up. Biol Psychiatry. 2009;66(11):1023-30.
[39] Simon AE, Grädel M, Cattapan-Ludewig K, Gruber K, Ballinari P, Roth B, Umbricht D.
Cognitive functioning in at-risk mental states for psychosis and 2-year clinical outcome.
Schizophr Res. 2012;142(1-3):108-15.
[40] Ruhrmann S, Schultze-Lutter F, Salokangas RK, Heinimaa M, Linszen D, Dingemans P,
Birchwood M, Patterson P, Juckel G, Heinz A, Morrison A, Lewis S, von Reventlow HG,
Klosterkötter J. Prediction of psychosis in adolescents and young adults at high risk:
results from the prospective European prediction of psychosis study. Arch Gen
Psychiatry. 2010;67(3):241-51.
[41] Salokangas RK, Ruhrmann S, von Reventlow HG, Heinimaa M, Svirskis T, From T,
Luutonen S, Juckel G, Linszen D, Dingemans P, Birchwood M, Patterson P, Schultze-
Lutter F, Klosterkötter J; EPOS group. Axis I diagnoses and transition to psychosis in
clinical high-risk patients EPOS project: prospective follow-up of 245 clinical high-risk
outpatients in four countries. Schizophr Res. 2012;138(2-3):192-7.
[42] Amminger GP, Schäfer MR, Papageorgiou K, Klier CM, Cotton SM, Harrigan SM,
Mackinnon A, McGorry PD, Berger GE. Long-chain omega-3 fatty acids for indicated
prevention of psychotic disorders: a randomized, placebo-controlled trial. Arch Gen
Psychiatry. 2010;67(2):146-54.
[43] Fusar-Poli P, Hobson R, Raduelli M, Balottin U. Reliability and validity of the
Comprehensive Assessment of the At Risk Mental State, Italian version (CAARMS-I).
Curr Pharm Des. 2012;18(4):386-91.
[44] Lemos-Giráldez S, Vallina-Fernández O, Fernández-Iglesias P, Vallejo-Seco G,
Fonseca-Pedrero E, Paíno-Piñeiro M, Sierra-Baigrie S, García-Pelayo P, Pedrejón-
Molino C, Alonso-Bada S, Gutiérrez-Pérez A, Ortega-Ferrández JA. Symptomatic and
functional outcome in youth at ultra-high risk for psychosis: a longitudinal study.
Schizophr Res. 2009;115(2-3):121-9.
[45] Kiss I, Kelemen O, Kéri S. Decreased peripheral expression of neuregulin 1 in high-risk
individuals who later converted to psychosis. Schizophr Res. 2012;135(1-3):198-9.
[46] Kotlicka-Antczak M, Pawełczyk T, Rabe-Jabłońska J, Pawełczyk A. PORT (Programme

of Recognition and Therapy): the first Polish recognition and treatment programme for
patients with an at-risk mental state. Early Interv Psychiatry doi: 10.1111/eip.12146.
[47] Lam MM, Hung SF, Chen EY. Transition to psychosis: 6-month follow-up of a Chinese
high-risk group in Hong Kong. Aust N Z J Psychiatry. 2006;40(5):414-20.
[48] Lee J, Rekhi G, Mitter N, Bong YL, Kraus MS, Lam M, Rapisarda A, Lee TS,
Subramaniam M, Chong SA, Keefe RS. The Longitudinal Youth at Risk Study (LYRIKS)-
-an Asian UHR perspective. Schizophr Res. 2013;151(1-3):279-83.
[49] Katsura M, Ohmuro N, Obara C, Kikuchi T, Ito F, Miyakoshi T, Matsuoka H, Matsumoto
K. A naturalistic longitudinal study of at-risk mental state with a 2.4year follow-up at a
specialized clinic setting in Japan. Schizophr Res doi: 10.1016/j.schres.2014.06.013.
[50] Koike S, Takano Y, Iwashiro N, Satomura Y, Suga M, Nagai T, Natsubori T, Tada M,
Nishimura Y, Yamasaki S, Takizawa R, Yahata N, Araki T, Yamasue H, Kasai K. A
multimodal approach to investigate biomarkers for psychosis in a clinical setting: the
integrative neuroimaging studies in schizophrenia targeting for early intervention and
prevention (IN-STEP) project. Schizophr Res. 2013;143(1):116-24.
[51] Zhang T, Li H, Woodberry KA, Seidman LJ, Zheng L, Li H, Zhao S, Tang Y, Guo Q, Lu
X, Zhuo K, Qian Z, Chow A, Li C, Jiang K, Xiao Z, Wang J. Prodromal psychosis
detection in a counseling center population in China: an epidemiological and clinical
study. Schizophr Res. 2014;152(2-3):391-9.
[52] Kim E, Jang JH, Park HY, Shim G, Hwang JY, Kim SN, Kwon JS. Pharmacotherapy and
clinical characteristics of ultra-high-risk for psychosis according to conversion status: a
naturalistic observational study. Early Interv Psychiatry. 2012;6(1):30-7.
[53] McGlashan T, Walsh B, Woods S. The Psychosis-Risk Syndrome. Handbook for
Diagnosis and Follow-Up. New York, NY: Oxford University Press. 2010.
[54] Yung AR, Yuen HP, McGorry PD, Phillips LJ, Kelly D, Dell'Olio M, Francey SM,
Cosgrave EM, Killackey E, Stanford C, Godfrey K, Buckby J. Mapping the onset of
psychosis: the Comprehensive Assessment of At-Risk Mental States. Aust N Z J
Psychiatry. 2005;39(11-12):964-71.
[55] Yung AR, Phillips LJ, Simmons MB, Ward J, Thompson P, French P, McGorry P.
CAARMS. Comprehensive Assessment of at Risk Mental States. Parkville Victoria: The
PACE Clinic, ORYGEN Research Centre, University of Melbourne, Department of
Psychiatry. 2006.

Eur Psychiatry (2015) 30a

Uploaded by

Copyright:

Available Formats

Eur Psychiatry (2015) 30a

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Eur Psychiatry (2015) 30a

Uploaded by

Copyright:

Available Formats

European Psychiatry 30 (2015) 405–416

Contents lists available at ScienceDirect

EPA guidance on the early detection of clinical high risk states of

In psychiatry, as in medicine, strenuous efforts are made to

Scale Onset Frequency Substance-use, co-morbidities Functioning

Attenuated psychotic symptoms (APS)

Brief limited intermittent, i.e. transient psychotic symptoms (BLIPS)

SOFAS: Social and Occupational Functioning Assessment Scale [3].

CAARMS 2006 version: Modelled on earlier versions of CAARMS [130]

BSIP: Basel Screening Modelled on the BPRS [82]

in APS samples was not available. Data on conversion rates in BLIPS

Fig. 4. Conversion rates at different follow-ups in samples meeting the ‘Cognitive

3.5.1. Single and combined CHR criteria 4. Recommendations

4.2. Proposed recommendations of the European Guidance Project 4.2.4. Recommendation 4

Frauke Schultze-Lutter, Chantal Michel, Stefanie J. Schmidt, Benno G. Schimmelmann,

Ruhrmann, Joachim Klosterkötter

S.1. Procedures and formulae of meta-analyses ..................................................................... 2

S.1. Procedures and formulae of meta-analyses

meta-analyses of univariate studies when the assessment of measure of interest is

episode psychosis is generally diagnosed according to DSM-IV, we used the proportions of

when z  2.58 [1].

as the related 95% CIs and z-values were computed.

S.2. Sensitivity analyses

analyses (see S.1.) were performed with these subgroups in addition.

percentages of conversion rates (𝐸 / 𝐸 ∗  100%) and compared for significant differences

using exploratory one-dimensional 2-tests (with df = number of effect sizes – 1) unadjusted

for multiple testing.

S.3. Supplementary Table 1

Observational study Co-morbidities: 47.6% with

Observational study Co-morbidities: Not reported

COPER: n=13 (31.7%)

Observational study Co-morbidities: 24.7% mood or

S.4. Supplementary Table 2

STable 2 Effect sizes (Ei, 𝐸 or 𝐸 ∗ ) at different follow-ups tx in samples meeting COPER,

STable 2 cont. (1)

Study tx N Ei, 𝑬 or 𝑬∗ 95% CIs Q (df) I2 Z

STable 2 cont. (2)

Study tx N Ei, 𝑬 or 𝑬∗ 95% CIs Q (df) I2 Z

Study tx N Ei, 𝑬 or 𝑬∗ 95% CIs Q (df) I2 Z

Study tx N Ei, 𝑬 or 𝑬∗ 95% CIs Q (df) I2 Z

S.5. Supplementary Table 3

STable 3 Pairwise comparison of conversion rates at different follow-ups tx in samples

S.6. Supplementary Table 4

STable 4 Pairwise comparison of conversion rates at different follow-ups tx in UHR

S.7. Supplementary Table 5

STable 5 Pairwise comparison of conversion rates at different follow-ups tx in UHR

° p  0.10; * p  0.05; ** p<0.01; *** p<0.001

S.8. References of Supplementary Material

[1] Eisend M. Metaanalyse. Meiring: Rainer Hampp Verlag. 2014.

multisite approach to prodromal schizophrenia research. Schizophr Bull. 2007;33(3):665-

[25] Schultze-Lutter F, Klosterkötter J, Picker H, Steinmeyer EM, Ruhrmann S. Predicting

[36] Manninen M, Lindgren M, Therman S, Huttunen M, Ebeling H, Moilanen I, Suvisaari J.

[46] Kotlicka-Antczak M, Pawełczyk T, Rabe-Jabłońska J, Pawełczyk A. PORT (Programme

You might also like

° p  0.10; * p  0.05; p<0.01; * p<0.001