Intervention Review

You have free access to this content

Atypical antipsychotics for psychosis in adolescents

  1. Ajit Kumar1,
  2. Soumitra S Datta2,3,*,
  3. Stephen D Wright4,
  4. Vivek A Furtado5,
  5. Paul S Russell6

Editorial Group: Cochrane Schizophrenia Group

Published Online: 15 OCT 2013

Assessed as up-to-date: 18 JUL 2012

DOI: 10.1002/14651858.CD009582.pub2


How to Cite

Kumar A, Datta SS, Wright SD, Furtado VA, Russell PS. Atypical antipsychotics for psychosis in adolescents. Cochrane Database of Systematic Reviews 2013, Issue 10. Art. No.: CD009582. DOI: 10.1002/14651858.CD009582.pub2.

Author Information

  1. 1

    Leeds and York Partnership NHS Foundation Trust, Psychiatry, Leeds, UK

  2. 2

    Institute of Psychiatry, King's College London, Child & Adolescent Psychiatry, London, UK

  3. 3

    Tata Medical Centre, Department of Palliative Care & Psycho-oncology, Kolkata, West Bengal, India

  4. 4

    Leeds Partnerships NHS Foundation Trust, Central & North West Community Mental Health Team, Leeds, West Yorkshire, UK

  5. 5

    Institute of Mental Health, Forensic Psychiatry, Nottingham, Nottinghamshire, UK

  6. 6

    Christian Medical College, Child & Adolescent Psychiatry, Vellore, Tamil Nadu, India

*Soumitra S Datta, Child & Adolescent Psychiatry, Institute of Psychiatry, King's College London, De Crespigny Park, London, SE5 8AF, UK. ssdatta2000@yahoo.com. ssdatta@doctors.org.uk.

Publication History

  1. Publication Status: New
  2. Published Online: 15 OCT 2013

SEARCH

 
Characteristics of included studies [ordered by study ID]
Aranda 2007

MethodsAllocation: randomised.
Blindness: open.
Duration: 6 months.
Design: single centre.
Country: Spain.


ParticipantsDiagnosis: psychosis (DSM-IV).
N = 50.
Age, years: range 12 to 18, mean age for quetiapine group 16.3 and olanzapine group 15.6.
Sex: not stated.
Setting: inpatient.


Interventions1. Quetiapine mean dose 532.8 mg/d. N = 24.

2. Olanzapine mean dose 9.7 mg/d. N = 26.


OutcomesLeaving the study early.

Unable to use:

Global state: CGI-S (no SD).

Mental state: PANSS (no SD).

Neuropsychological test for cognition: WCST (no usable data).

Adverse effects (no usable data).


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described in the report.

Allocation concealment (selection bias)Unclear riskNot described in the report.

Blinding of participants and personnel (performance bias)
All outcomes
High riskThis was an open-label study.

Blinding of outcome assessment (detection bias)
All outcomes
High riskThis was an open-label study.

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskThe issue was partially addressed.

Selective reporting (reporting bias)Unclear riskThe study reported multiple outcomes but partially (e.g. no SD for mean end point PANSS score).

Other biasUnclear riskThis seems to be a well-designed study, but not all results are reported.

DelBello 2008

MethodsAllocation: randomised.
Blindness: open label.
Duration: 3 weeks.
Design: multi-centre.
Country: USA.


ParticipantsDiagnosis: schizophrenia and schizoaffective disorder (DSM-IV-TR).
N = 17.
Age, years: range 10 to 17, mean age 14.6.
Sex: 11 male, 6 female.
Inclusion criteria: diagnosis of schizophrenia or schizoaffective disorder as defined by the DSM-IV-TR; diagnosis confirmed by KID-SCID; BPRS score ≥ 35 with a score ≥ 4 on at least one of the items—unusual thought content, suspiciousness, hallucinations or conceptual disorganisation. Only participants with a body mass index (BMI) between the 5th and 95th percentiles were included.
Exclusion criteria: currently on stable well-tolerated treatment; suspected or established substance-induced psychotic disorder; treatment with clozapine in the last 12 weeks; a depot antipsychotic in the last 4 weeks or a monoamine oxidase inhibitor (MAOI) in the past 2 weeks; imminent risk of suicide or homicide; IQ ≤ 70; autism or other pervasive developmental disorder; pregnancy, breast-feeding or unwillingness to use contraceptions; any serious medical or neurological illness; any screening laboratory value that deviated significantly from the reference range; history of cardiac problems, QTc prolongation ≥ 460 ms; or DSM-IV-TR-defined psychoactive substance abuse or dependence within the preceding month.
Setting: unclear.


Interventions1. Ziprasidone 80 mg/d. N = 8.

2. Ziprasidone 160 mg/d. N = 9.


OutcomesLeaving the study.

Global state: CGI-S.

Mental state: BPRS-A.

Unable to use:

Adverse effects: SARS, AIMS, BARS (no usable data).

Laboratory tests (no usable data).


NotesThe study was divided into two periods: fixed doses of ziprasidone—three weeks; flexible doses—24 weeks. Only outcome data after a fixed-dose trial were usable. The study was funded by Pfizer Inc. The authors reported separate data for bipolar disorder and schizophrenia/schizoaffective disorder.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskDetails of sequence generation were not mentioned. Because the study population had a narrow age range, the different groups were comparable.

Allocation concealment (selection bias)Unclear riskNot described in the report.

Blinding of participants and personnel (performance bias)
All outcomes
High riskThis was an open-label trial.

Blinding of outcome assessment (detection bias)
All outcomes
High riskThis was an open-label trial.

Incomplete outcome data (attrition bias)
All outcomes
Low riskAuthors reported the number of participants who had to discontinue with the study. The primary aim of the study was to test the tolerability of high- vs low-dose regimens of ziprasidone. Measuring efficacy was not the primary aim of this study. Adverse effects experienced by study participants were reported in sufficient detail.

Selective reporting (reporting bias)Low riskAuthors reported many outcome measures, especially adverse effects, in sufficient detail.

Other biasUnclear riskStudy participants were selected from 10 centres and met fairly broad inclusion criteria. This could lead to inconclusive evidence of a particular condition. Authors reported findings separately for bipolar disorder and schizophrenia/schizoaffective disorder.

Findling 2008

MethodsAllocation: randomly assigned.
Blindness: double.
Duration: 6 weeks.
Design: multi-centre.
Country: USA, Europe, South America, Asia, the Carribean, South Africa.


ParticipantsDiagnosis: schizophrenia (DSM-IV).
N = 302.
Age, years: range 13 to 17, mean age 15.4 + 1.4.
Sex: 171 male, 131 female.
Inclusion criteria: diagnosis of schizophrenia as defined by the DSM-IV; diagnosis of schizophrenia confirmed by an adequately trained clinician (e.g. child psychiatrist) by means of the K-SADS-PL, PANSS score ≥ 70.
Exclusion criteria: psychiatric comorbidity requiring pharmacotherapy, any evidence of suicide risk. Current or past history of schizoaffective disorder, major depression, mental retardation, neuroleptic malignant syndrome, any neurological disorder except Tourette's syndrome, severe head trauma or unstable medical disorder. Participants with resistant illness to two different antipsychotic drugs, sexually active adolescent boys or girls who did not agree to use contraceptives, positive screens for illegal drugs within 3 months of baseline.
Setting: inpatient and outpatient.


Interventions1. Placebo. N = 100.

2. Aripiprazole 10 mg/d. N = 100.

3. Aripiprazole 30 mg/d. N = 102.


OutcomesLeaving the study.

Global state: CGI-S, CGI-I, CGAS.

Mental state: PANSS.

Quality of life: P-QLES-Q.

Adverse effects: AIMS, BARS, SAS.

Biochemistry.

Unable to use:

Vital signs (no usable data).

Electrocardiogram (ECG) (no usable data).


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskThe randomisation was stratified and was done separately across three geographical categories.

Allocation concealment (selection bias)Unclear riskNot described in the report.

Blinding of participants and personnel (performance bias)
All outcomes
Low riskThis was a double-blind placebo-controlled trial.

Blinding of outcome assessment (detection bias)
All outcomes
Low riskThis was a double-blind placebo-controlled trial.

Incomplete outcome data (attrition bias)
All outcomes
Low riskIncomplete outcome data were analysed.

Selective reporting (reporting bias)Low riskAuthors reported multiple outcome measures.

Other biasLow riskThis was a well-designed RCT.

Haas 2009

MethodsAllocation: randomised.
Blindness: double.
Duration: 8 weeks.
Design: multi-centre.
Country: Belgium, Bulgaria, Czech Rep, Estonia, Germany, Poland, Romania, USA.


ParticipantsDiagnosis: catatonic, disorganised, paranoid, residual, undifferentiated schizophrenia (DSM-IV), acute episode.
N = 257.
Age, years: range 13 to 17, mean age 15.6.
Sex: 145 male, 112 female.
Inclusion criteria: diagnosis of schizophrenia as defined by the DSM-IV, currently hospitalised with an acute episode (PANSS score 60 to 120). Negative serum pregnancy test.
Exclusion criteria: schizophreniform disorder, significant suicidal risk or risk of violence, past history of neuroleptic malignant syndrome, tardive dyskinesia, known or suspected seizure disorder BMI ≤ 5th percentile or BMI > 95th percentile and administration of more than two doses of drug in the drug-free washout period.
Setting: inpatient.


Interventions1. Risperidone dose range 1.5 to 6.0 mg/d. N = 125.

2. Risperidone dose range 0.15 to 0.6 mg/d. N = 132.

In association with psychotherapy and psychoeducation.


OutcomesLeaving the study.

Global state: CGI-S, CGI-I.

Mental state: PANSS.

Adverse effects: SAS, AIMS, BARS.

Unable to use:

Laboratory tests (no usable data).


NotesInvestigators were required as per protocol to adjust medications up to the maximum tolerated dose over a period of 12 days to ensure that a full dose range would be explored for safety. The dose was to remain stable during the last 4 weeks of the double-blind period.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskComputer-generated schedule.

Allocation concealment (selection bias)Unclear riskNot described in the report.

Blinding of participants and personnel (performance bias)
All outcomes
Low riskThis was a double-blind study.

Blinding of outcome assessment (detection bias)
All outcomes
Low riskThis was a double-blind study.

Incomplete outcome data (attrition bias)
All outcomes
Low riskIncomplete data were accounted for by using intention-to-treat analysis.

Selective reporting (reporting bias)Low riskSupplementary data in journal website gives details of many of the outcome measures.

Other biasLow riskThis was a well-designed RCT.

Huo 2007

MethodsAllocation: randomised.
Blindness: unclear.
Duration: 8 weeks.
Design: single centre.
Country: China.


ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 40.
Age, years: mean age 14.
Sex: 21 male, 19 female.
Inclusion criteria: diagnosis of schizophrenia, age ≤ 16 years.
Exclusion criteria: patients with organic diseases.
Setting: inpatient and outpatient.


Interventions1. Risperidone mean dose 3.18 ± 0.66 mg/d, range 2 to 4 mg/d. N = 20.

2. Perphenazine mean dose 16.2 ± 6.4 mg/d, range 10 to 24 mg/d. N = 20.


OutcomesMental state: BPRS.
Adverse effects: TESS.


NotesAverage length of illness: 2 to 8 months.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described in the report.

Allocation concealment (selection bias)Unclear riskNot described in the report.

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskNot described in the report.

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot described in the report.

Incomplete outcome data (attrition bias)
All outcomes
Low riskThe results accounted for all participants who started the study.

Selective reporting (reporting bias)Low riskThe study reported all outcomes it had intended to report at the beginning in the study.

Other biasLow riskAlthough the study was in Chinese, we found after translation that reporting was done in a clear and easy to understand way.

Jensen 2008

MethodsAllocation: randomised.
Blindness: open.
Duration: 12 weeks.
Design: single centre.
Country: USA.


ParticipantsDiagnosis: schizophrenia, schizoaffective disorder, schizophreniform disorder, psychotic disorder NOS (DSM-IV).
N = 30.
Age, years: range 10 to 18, mean age 15.2 ± 2.1.
Sex: 20 male, 10 female.
Inclusion criteria: boys and girls aged 10 to 18 years, inclusive, with a diagnosis of schizophrenia/schizoaffective disorder, schizophreniform disorder, or psychotic disorder NOS. At least one positive or negative symptom associated with schizophrenia of moderate or greater severity on PANSS that had been present throughout the past 2 weeks.
Exclusion criteria: diagnosis of mental retardation, affective disorder (i.e. major depressive disorder or bipolar disorder) with psychotic features, current alcohol or drug dependence or abuse, history of serious adverse reactions or non-response to an adequate trial of any of the proposed treatments, pregnant or refused to practice contraception, serious and unstable medical condition.
Setting: inpatient and outpatient.


Interventions1. Risperidone mean dose 3.4 ± 1.5 mg/d, range 1 to 6 mg/d. N = 10.

2. Olanzapine mean dose 14.6 ± 4.6 mg/d, range 5 to 20 mg/d. N = 10.

3. Quetiapine mean dose 611 ± 253.4 mg/d, range 100 to 800 mg/d. N = 10.


OutcomesLeaving the study early.

Global state: CGI-S, CGAS.

Mental state: PANSS.

Adverse effects: AIMS, SAS.

Unable to use:

Laboratory tests (no usable data).


NotesOf 30 participants 27 were inpatients.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandom sequence was performed using computer-generated randomisation schedule.

Allocation concealment (selection bias)Unclear riskNot described in the report.

Blinding of participants and personnel (performance bias)
All outcomes
High riskThis was an open-label study.

Blinding of outcome assessment (detection bias)
All outcomes
High riskThis was an open-label study.

Incomplete outcome data (attrition bias)
All outcomes
Low riskIntention-to-treat analysis was done using the 'last observation carried forward' method for participants who did not complete the study.

Selective reporting (reporting bias)Low riskOutcomes were reported in sufficient detail.

Other biasHigh riskAs the study was open label, it was probably subjected to interviewer's bias or researcher's bias.

Kryzhanovskaya 2009

MethodsAllocation: randomised.
Blindness: double.
Duration: 6 weeks.
Design: multi-centre.
Country: USA, Russia.


ParticipantsDiagnosis: paranoid, disorganised, catatonic, undifferentiated, residual schizophrenia (DSM-IV-TR).
N = 107.
Age, years: range 13 to 17, mean age 16.
Sex: 75 male, 32 female.
Inclusion criteria: diagnosis confirmed by the K-SADS-PL. Baseline score ≥ 35 on the anchored version of the BPRS-C (21 items), with a score of 3 or higher on at least one of the following BPRS-C items at enrolment and randomisation: hallucinations, delusions or peculiar fantasies.
Exclusion criteria: previous participation in a clinical trial of oral olanzapine. Treatment within 30 days of the trial with a drug without regulatory approval for any indication. Previous non-response to an adequate dose/duration of olanzapine treatment. Pregnancy, nursing, or refusal to practise contraception (for females). Acute or unstable medical condition. Prolactin levels > 200 ng/L at randomisation. DSM-IV-TR substance dependence within 30 days. Current diagnosis of a comorbid psychiatric or developmental disorder.
Setting: inpatient and outpatient.


Interventions1. Olanzapine mean dose 11.1 ± 4.0 mg/d, range 2.5 to 20 mg/d. N = 72.

2. Placebo. N = 35.


OutcomesLeaving the study early.

Global state: CGI-I, CGI-S.

Mental state: BPRS-C, PANSS, OAS.

Adverse effects: SAS, BARS, AIMS.

Unable to use:

Laboratory tests (no usable data).


NotesMean age at onset of illness ∽ 13 years.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described in the report.

Allocation concealment (selection bias)Unclear riskNot described in the report.

Blinding of participants and personnel (performance bias)
All outcomes
Low riskThere was a double-blind phase followed by an open-label phase.

Blinding of outcome assessment (detection bias)
All outcomes
Low riskThere was a double-blind phase followed by an open-label phase.

Incomplete outcome data (attrition bias)
All outcomes
Low riskAppropriate intention-to-treat analysis was performed.

Selective reporting (reporting bias)Low riskMultiple outcome measures were reported in detail.

Other biasLow riskThis was a well-designed RCT.

Kumra 1996

MethodsAllocation: randomised.
Blindness: double.
Duration: 6 weeks
Design: single centre.
Country: USA.


ParticipantsDiagnosis: disorganised, undifferentiated, paranoid schizophrenia (DSM-III-R).
N = 21.
Age, years: range 6 to 18, mean 14 ± 2.3 .
Sex: 11 male, 10 female.
Inclusion criteria: diagnosis of schizophrenia as defined by the DSM-III-R, with documented psychotic symptoms by the age of 12 years; intolerance, non-response or both to at least two different neuroleptic drugs; IQ of 70 or greater. In other words, this study included only adolescents with treatment-resistant schizophrenia.
Exclusion criteria: neurological or medical disease.
Setting: inpatient.


Interventions1. Clozapine mean dose 176 ± 149 mg/d, range 25 to 525 mg/d. N = 10.

2. Haloperidol mean dose 16 ± 8 mg/d, range 7 to 27 mg/d. N = 11.


OutcomesLeaving the study early.

Global state: CGI, CGAS.

Mental state: BPRS, SANS, SAPS, B-HPRS).

Adverse effects: SAS, TESS, AIMS.


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandomisation sequence was done using table of random numbers.

Allocation concealment (selection bias)Unclear riskNot described in the report.

Blinding of participants and personnel (performance bias)
All outcomes
Low riskRaters, treating physicians and nurses were blind to the study status.

Blinding of outcome assessment (detection bias)
All outcomes
Low riskRaters were blind to the study status.

Incomplete outcome data (attrition bias)
All outcomes
Low riskAppropriate intention-to-treat analysis was conducted.

Selective reporting (reporting bias)High riskPositive findings were highlighted in the 'Results' section. The high incidence of neutropenia in the clozapine group has been reported but not adequately discussed.

Other biasHigh riskThis is a well-designed RCT for treatment-resistant schizophrenia in childhood, but the total number of children recruited is small.

Kumra 2008

MethodsAllocation: randomised.
Blindness: double.
Duration: 12 weeks.
Design: multi-centre.
Country: USA.


ParticipantsDiagnosis: treatment-resistant schizophrenia, schizoaffective disorder (DSM-IV).
N = 39.
Age, years: range 10 to 18, mean age 15.6 (2.1).
Sex: 21 male, 18 female.
Inclusion criteria: diagnosis of schizophrenia or schizoaffective disorder based on a structured interview, non-response to at least two different neuroleptic drugs, baseline BPRS ≥ 35. In other words, this study included only adolescents with treatment-resistant schizophrenia.
Exclusion criteria: premorbid diagnosis of mental retardation (IQ < 70); history of serious adverse reactions to proposed treatments; pregnancy; serious or unstable medical condition; failure to respond to adequate trials of clozapine (≥ 300 mg/d for 12 weeks) or olanzapine (≥ 20 mg/d for 8 weeks).
Setting: inpatient.


Interventions1. Clozapine mean dose 403.1 ± 201.8 mg/d, range 50 to 700 mg/d. N = 18.

2. "High-dose" olanzapine mean dose 26.2 ± 6.5 mg/d, range 10 to 30 mg/d. N = 21.


OutcomesLeaving the study early.

Global state: CGI, CGAS.

Mental state: BPRS, SANS.

Adverse effects: TESS, AIMS.

Laboratory tests.


NotesAge of onset of psychosis, years: 12.7 ± 2.4 (clozapine group); 11.75 ± 3.2 (olanzapine group).


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskComputer-generated randomisation sequence was followed.

Allocation concealment (selection bias)Low riskAdequate attempts were made to maintain allocation concealment (using numbered containers, centralised telephone, etc).

Blinding of participants and personnel (performance bias)
All outcomes
Low riskThis was a double-blind RCT, but the details of blinding and how it was maintained had not been described in the report.

Blinding of outcome assessment (detection bias)
All outcomes
Low riskThis was a double-blind RCT, but the details of blinding and how it was maintained had not been described in the report.

Incomplete outcome data (attrition bias)
All outcomes
Low riskAppropriate intention-to-treat analysis was performed. All participants who had been randomly assigned were included. in the analysis

Selective reporting (reporting bias)Low riskAdequate numbers of outcome measures were reported.

Other biasLow riskThis was a well-designed RCT.

Sikich 2004

MethodsAllocation: randomised.
Blindness: double.
Duration: 8 weeks.
Design: multi-centre.
Country: North America.


ParticipantsDiagnosis: psychotic disorder (K-SADS-PL,SCID).
N = 51.
Age, years: range 8 to 19; mean age 14 years 8 months.
Sex: 30 male, 20 female.
Inclusion criteria: at least one positive psychotic symptom of moderate or greater severity on the BPRS-C, present for two weeks, IQ > 69.
Exclusions: acute substance intoxication or withdrawal; history of adverse reactions or non-response to study medications; prior diagnosis of a pervasive developmental disorder; serious medical illness; pregnancy or refusal to practise contraception; imminent risk of harm to self or others.
Setting: inpatient and outpatient.


Interventions1. Risperidone mean dose 3.3 mg/d. N = 19.

2. Olanzapine mean dose 12.3 mg/d. N = 16.

3. Haloperidol mean dose 5.3 mg/d. N = 15.


OutcomesLeaving the study early.

Global state: CGI, CGI-S.

Mental state : BPRS, CPRS.

Adverse effects: AIMS, SAS.

Unable to use:

Laboratory tests (no usable data).


Notes36% discontinuation rate.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskComputer-generated randomisation schedule, stratified by age.

Allocation concealment (selection bias)Unclear riskNot adequately described in the study.

Blinding of participants and personnel (performance bias)
All outcomes
Low riskThis was a double-blind study, but the details of how blinding was maintained were not reported.

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskAs the study compares drugs from different classes, the raters could have made a guess about young people being on haloperidol.

Incomplete outcome data (attrition bias)
All outcomes
Low riskAll participants who were randomly assigned were included in the analysis.

Selective reporting (reporting bias)Unclear riskMany different outcome measures were reported.

Other biasHigh riskThe analysis did not use correction techniques to take into account multiple analyses on a small sample. The authors justified this by stating that these analyses were 'exploratory' rather than 'inferential'.

Sikich 2008

MethodsAllocation: randomised.
Blindness: double.
Duration: 8 weeks.
Design: multi-centre.
Country: North America.


ParticipantsDiagnosis: schizophrenia, schizoaffective disorder or schizophreniform disorder (DSM-IV).
N = 116.
Age, years: 8 to 19.
Sex: 75 male, 41 female.
Inclusion criteria: current positive psychotic symptoms of at least moderate intensity on PANSS.
Exclusions: prior evidence of mental retardation; current major depressive episode; active substance abuse; acute substance intoxication or withdrawal; history of intolerance or non-response to study medications; history of an adequate trial of any study medications; imminent risk of harm to self or others
Setting: inpatient and outpatient.


Interventions1. Molindone mean dose 59.9 ± 33.5 mg/d. N = 40.

2. Olanzapine mean dose 11.4 ± 5.0 mg/d. N = 35.

3. Risperidone mean dose 2.8 ± 1.4 mg/d. N = 41.


OutcomesLeaving the study early.

Global state: CGI, Adult and Child Functional Assessment Scale.

Mental state: PANSS, BPRS-C.

Adverse effects: SARS, BARS, AIMS.

Laboratory tests.


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot adequately described in the study.

Allocation concealment (selection bias)Unclear riskNot adequately described in the study.

Blinding of participants and personnel (performance bias)
All outcomes
Low riskThis was a double-blind study, and blinding was maintained even after discontinuation of one of the arms (olanzapine group) of the study for ethical reasons.

Blinding of outcome assessment (detection bias)
All outcomes
Low riskBlinding was maintained even after discontinuation of one of the arms of the study.

Incomplete outcome data (attrition bias)
All outcomes
Low riskAll participants who had been randomly assigned were included in the final analysis.

Selective reporting (reporting bias)Low riskMultiple outcome measures are described in sufficient detail.

Other biasLow riskThis is a well-designed RCT.

Swadi 2010

MethodsAllocation: randomised.
Blindness: single blind (open label with blind midpoint and end point assessments).
Duration: 6 weeks.
Design: single centre.
Country: New Zealand.


ParticipantsDiagnosis: first-onset psychotic disorder or mood disorder with psychotic features (DSM-IV).
N = 22.
Age, years: range 15 to 18.
Sex: 13 male, 9 female.
Inclusion criteria: individuals younger than 19 years of age, with a first-onset psychotic disorder or a mood disorder with psychotic features according to DSM-IV.
Exclusion criteria: individuals with alcohol or substance dependence not in full remission and persons who had received earlier treatment with atypical antipsychotic drugs.
Setting: inpatient.


Interventions1. Risperidone mean dose 2.9 mg/d, range 1.5 to 5 mg/d. N = 11.

2. Quetiapine mean dose 607 mg/d, range 100 to 800 mg/d. N = 11.


OutcomesGlobal state: CGI-S.
Mental state: PANSS, BPRS, HAM-D, YMRS.
Adverse effects: AIMS, SARS, BARS.
Laboratory tests.


NotesThis is a well-designed study involving a small number of participants from a single centre.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskStratified randomisation process, four strata.

Allocation concealment (selection bias)Unclear riskNot adequately described in the study.

Blinding of participants and personnel (performance bias)
All outcomes
High riskOpen-label study with blind midpoint and end point assessments.

Blinding of outcome assessment (detection bias)
All outcomes
High riskOpen-label study with blind midpoint and end point assessments.

Incomplete outcome data (attrition bias)
All outcomes
Low riskClear flow chart for study patients accounting for all participants.

Selective reporting (reporting bias)Low riskNo source of selective reporting could be found.

Other biasUnclear riskLack of double blindness puts the study under suspicion of further biases.

Xiong 2004

MethodsAllocation: randomised.
Blindness: not clear.
Duration: 8 weeks.
Design: single centre.
Country: China.


ParticipantsDiagnosis: childhood-onset schizophrenia (CCMD-2-R).
N = 60.
Age, years: 7 to 16, mean age ∽ 14.
Sex: 34 male, 26 female.
Inclusion criteria: children with the diagnosis of schizophrenia according to CCMD-2-R should be between 7 and 16 years of age with no physical problems and no organic neurological disease.
Setting: inpatient.


Interventions1. Risperidone dose range 0.5 to 5 mg/d. N = 30.

2. Chlorpromazine dose range 50 to 400 mg/d. N = 30.


OutcomesMental state: BPRS.
Adverse effects: TESS.


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot adequately described in the study.

Allocation concealment (selection bias)Unclear riskNot adequately described in the study.

Blinding of participants and personnel (performance bias)
All outcomes
High riskBlinding not clear.

Blinding of outcome assessment (detection bias)
All outcomes
High riskBlinding not clear.

Incomplete outcome data (attrition bias)
All outcomes
Low riskData reported for all participants.

Selective reporting (reporting bias)Low riskThe study reported all of the outcomes that it had intended to report at the start of study.

Other biasUnclear riskUnclear blinding makes this study prone to further biases.

 
Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

Amminger 2006Allocation: randomised.
Participants: individuals (13 to 24 years) at ultra-high risk (UHR) for psychosis
Interventions: omega-3 fatty acids vs placebo.
Outcomes: no separate data reported for 13 to 17 years age group.

Antropov 1981Allocation: not randomised, open-label study.

Berger 2007Allocation: randomised.
Participants: people in first psychotic episode, drug naive, age range 15 to 25 years.
Interventions: 200 mg/d vs 400 mg/d quetiapine fumarate.
Outcomes: no separate data reported for 15 to 17 years age group.

Bertelsen 2005Allocation: randomised.
Participants: people with first episode of schizophrenia or psychosis, age range 18 to 45 years.

Buchsbaum 2007Allocation: randomised.
Participants: people with a diagnosis of psychosis NOS (DSM-IV), age range 13 to 21 years.
Interventions: olanzapine vs haloperidol.
Outcomes: no usable data.

Chen 2007Allocation: randomised.
Participants: children and adolescents with first episode of schizophrenia, most participants younger than 13 years of age.

Davidson 2004Allocation: randomised.
Participants: people with first episode and early psychosis, no age range given.

Gao 2007Allocation: randomised.
Participants: children and adolescents with schizophrenia, most participants younger than 13 years of age.

Jenner 2004Allocation: not randomised, retrospective study.

Johnson 2004Allocation: randomised.
Participants: adolescents with diagnosis of schizophrenia (DSM-IV), age range 13 to 17 years.
Interventions: risperidone vs placebo.
Outcomes: no data provided, no usable data.

Killackey 2006Allocation: randomised.
Participants: people with first episode of psychosis, age range not clear.

Klier 2005Allocation: randomised.
Participants: people with diagnosis of "at-risk-mental-state" for psychosis, age range 13 to 25 years.
Interventions: omega-3 fatty acids and standard care vs standard care.
Outcomes: no usable data.

Leblanc 2006Allocation: randomised.
Participants: people with schizophrenia or related psychosis, age range not clear.

Leclerc 2006Allocation: randomised.
Participants: people with first psychotic episode, age range not clear.

Liang 2003Allocation: not randomised, blinding unclear.

Linszen 2006Allocation: randomised.
Participants: people with schizophrenia, mean age at onset of psychosis 19.3 years, focus on young adults rather than adolescents.

Lv 2004Allocation: randomised.
Participants: adolescents with schizophrenia.
Intervention: clozapine vs risperidone.
Outcomes: insufficient data, unable to use data.

Malik 1980Allocation: randomised.

Participants: adolescents with a diagnosis of acute, catatonic, paranoid or simple schizophrenia; age range 14 to 19 years (mean age 17 years).

Interventions: loxapine vs trifluoperazine.

Reason for exclusion: This study will be considered for the other review by the authors on typical antipsychotic medications. We excluded this study, done in the 1970s, from the current review, as this study did not provide any atypical antipsychotic medication in any of the arms.

Mathai 2004Allocation: randomised.
Participants: children with psychiatric disorders, no specific diagnosis, age range 4 to 14 years (mean age 9 years).

McConville 2003Allocation: not randomised, open-label study.

McGlashan 2003Allocation: randomised.
Participants: people meeting criteria of schizophrenia prodrome, most study participants not within 13 to 17 years age range.

McGorry 2007Allocation: randomised.
Participants: people meeting ultra-high risk (UHR) criteria for psychotic disorder, mean age 18.36 years, most study participants not within 13 to 17 years age range.

Newton 2005Allocation: not randomised, open-label study.

Otsuka 2005Allocation: randomised.
Participants: people with diagnosis of schizophrenia spectrum disorder or bipolar spectrum disorder, age range 13 to 17 years.
Interventions: aripiprazole.
Outcomes: a phase II study, insufficient data, unable to use data.

Power 2004Allocation: randomised.
Participants: people with first episode of psychosis, mean age 26.3 ± 6.2 years, most participants not within 13 to 17 years age range.

Schepp 1999Allocation: randomised.
Participants: people with a DSM-IV diagnosis of schizophrenia, age range 15 to 19 years.
Interventions: behavioural self-management study.
Outcomes: insufficient data, unable to use data.

Sela 2003Allocation: not randomised.

Stain 2006Allocation: randomised.
Participants: youths at risk for psychosis, age range not clear.

Tandon 2005Allocation: randomised.
Participants: hospitalised children and adolescents with a diagnosis of acute or transient psychotic disorder.
Interventions: risperidone and placebo.
Outcomes: no data comparing groups available from abstracts, no response from corresponding author to our request for full paper.

Ueland 2004Allocation: randomised.

Participants: individuals with a diagnosis of schizophrenia, schizoaffective disorder, schizotypal personality disorder, bipolar disorder, psychotic disorder NOS, major depressive disorder (DSM-IV); age range 12 to 18 years (mean 15.3 years).

Interventions: cognitive remediation programme (CRP) + psychoeducational treatment programme (PTP) vs PTP.

van Bruggen 2003Allocation: randomised.
Participants: adolescents and young adults with first or second episode of schizophrenia, schizophreniform or schizoaffective disorder (DSM-IV), age range 16 to 28 years.
Interventions: olanzapine vs risperidone.
Outcomes: no separate data reported for 16 to 17-year-olds. Corresponding author could not be contacted.

van Nimwegen 2006Allocation: randomised.
Participants: people with schizophrenia, schizophreniform or schizoaffective disorder (DSM-IV), mean age ∼ 25 years.

Versiani 1978Allocation: randomised.

Participants: adolescents with a diagnosis of schizophrenia (DSM-II), age range 13 to 18 years (mean age 16.1 years).

Interventions: loxapine vs haloperidol.

Reason for exclusion: This study will be considered by the authors for the other review on typical antipsychotic medications. We excluded this study, done in the 1970s, from the current review, as this study did not provide any atypical antipsychotic medication in any of the arms.

Wang 2007Allocation: randomised.
Participants: children and adolescents with schizophrenia, most participants younger than 13 years.

Wykes 2007Allocation: randomised.

Participants: adolescents and adults with a diagnosis of schizophrenia (DSM-IV), age range 14 to 22 years (mean age 18.2 years).

Interventions: cognitive remediation therapy (CRT) vs treatment as usual, details of treatment as usual not given.

Xiu 2004Allocation: randomised.
Participants: people with schizophrenia, mean age ∼ 27 years.

Yang 2007Allocation: randomised.
Participants: people with schizophrenia, mean age ∼ 11 years.

Yao 2003Allocation: randomised.
Participants: children with schizophrenia, mean age ∼ 11 years.

Yi 2006Allocation: randomised.
Participants: children and adolescents with schizophrenia, most study participants younger than 13 years.

Zhang 2007Allocation: randomised.
Participants: children with mean age 10.07 ± 2.35 years.

Zhou 2007Allocation: randomised.
Participants: children and adolescents with schizophrenia, most study participants younger than 13 years.

杨玲 2004Allocation: randomised.
Participants: children and adolescents with schizophrenia, age range 6 to 14 years.
Interventions: chlorpromazine vs clozapine vs risperidone.
Outcomes: no separate data given for 13 to 14 years age group.

谭友果 2002Allocation: not randomised, blinding not stated.

 
Characteristics of ongoing studies [ordered by study ID]
Alaghband-rad 2006

Trial name or titleManagement of first episode of psychoses in Iran: unique features and challenges.

MethodsRandomised.

ParticipantsDrug-naive inpatients, aged 15 to 60 years, who were diagnosed with first episode of psychosis .

InterventionsTreatment as usual (TAU) vs standard telephone follow-up (ST-TF) vs standard-home visit (ST-HV).

OutcomesComprehensive batteries of clinical ratings, cognitive and neuropsychological tests.

Starting date2006.

Contact informationrad@ams.ac.ir

NotesCorresponding author did not respond to our enquiry about the study

Another study under the same study ID includes adolescents (15 to 18 years) as participants. It seems highly likely that this is part of the main study, the characteristics of which are mentioned above.

AstraZeneca 2004

Trial name or titleA 6-week, multi-centre, randomised, double-blind, parallel-group, placebo-controlled, phase 3b study of the efficacy and safety of quetiapine fumarate (Seroquel™) immediate-release tablets compared with placebo in adolescents with schizophrenia.

MethodsRandomised.

ParticipantsMales and females aged 13 to 17 years with documented clinical diagnosis of schizophrenia.

InterventionsQuetiapine fumarate (Seroquel™) vs placebo.

OutcomesPANSS, level of functioning, safety, efficacy, tolerability, hostility, aggression.

Starting dateJuly 2004.

Contact informationClinicalTrials.gov identifier NCT00090324. AstraZeneca Information Center (8 AM to 7 PM EST). Tel 001-800-236-9933.

Notes

AstraZeneca 2005

Trial name or titleA 26-week, multi-centre, open-label, phase 3b study of the safety and tolerability of quetiapine fumarate (Seroquel™) immediate-release tablets in daily doses of 400 mg to 800 mg in children and adolescents with bipolar I disorder and in adolescents with schizophrenia.

MethodsRandomised.

ParticipantsMales and females aged 10 to 17 years with documented clinical diagnosis of schizophrenia or bipolar I disorder.

InterventionsQuetiapine fumarate (Seroquel™) 400 mg/d versus 800 mg/d.

OutcomesAdverse events (AEs); rate of participant withdrawal due to AEs; changes in clinical laboratory test results and SARS, BARS and AIM scores; changes in menses for female participants; changes in weight and BMI; changes in CGAS.

Starting dateJuly 2004.

Contact informationClinicalTrials.gov identifier NCT00227305. AstraZeneca Information Center (8 AM to 7 PM EST). Tel 001-800-236-9933.

Notes

Bechdolf 2007

Trial name or titleDevelopment and pilot evaluation of modified cognitive behavioural therapy for adolescents with early-onset psychosis.

MethodsRandomised.

ParticipantsMales and females aged 14 to 18 years with DSM-IV diagnosis of schizophrenia, schizophreniform or schizoaffective disorder, delusional disorder.

InterventionsModified cognitive behavioural therapy (mCBT) + treatment as usual (TAU) vs TAU.

OutcomesPANSS, social functioning (GAF), suicide, suicide attempts, rehospitalisation, severe depressive symptom exacerbation, quality of life.

Starting dateMay 2007.

Contact informationClinicalTrials.gov identifier NCT00465920. Andreas Bechdolf, PD, Dr; +49 221 478 3869; andreas.bechdolf@uk-koeln.de. Bettina Pohlmann, Dr; + 49 221 478 3870; bettina.pohlmann@uk-koeln.de.

NotesmCBT is an individual outpatient treatment consisting of 20 sessions and 5 psychoeducational sessions with parents.

Pfizer 2005

Trial name or titleSix-week, double-blind, placebo-controlled phase III trial evaluating the efficacy, safety and pharmacokinetics of flexible doses of oral ziprasidone in adolescent participants with schizophrenia.

MethodsRandomised.

ParticipantsMales and females aged 13 to 17 years with DSM-IV diagnosis of schizophrenia.

InterventionsZiprasidone vs placebo.

OutcomesBPRS, PANSS, CGI.

Starting dateApril 2006.

Contact informationClinicalTrials.gov Identifier: NCT00257192. United States: Food and Drug Administration.

Notes

 
Comparison 1. Atypical antipsychotics vs placebo (only short term)

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Global state: No response (CGI-S)1107Risk Ratio (M-H, Fixed, 95% CI)0.84 [0.65, 1.10]

 2 Mental state: 1. No response2304Risk Ratio (M-H, Fixed, 95% CI)0.76 [0.63, 0.92]

    2.1 No response (BPRS-C)
1107Risk Ratio (M-H, Fixed, 95% CI)0.84 [0.65, 1.10]

    2.2 Not achieving remission (PANSS)
1197Risk Ratio (M-H, Fixed, 95% CI)0.72 [0.56, 0.94]

 3 Mental state: 2. Change in PANSS score (data skewed, high score = good)Other dataNo numeric data

 4 Adverse effects: 1. Different adverse effects (binary measures)2Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    4.1 Cardiovascular—dizziness
1107Risk Ratio (M-H, Fixed, 95% CI)2.92 [0.37, 23.30]

    4.2 Central nervous system—somnolence
1107Risk Ratio (M-H, Fixed, 95% CI)8.26 [1.15, 59.61]

    4.3 Endocrine—clinically significant low prolactin
1302Risk Ratio (M-H, Fixed, 95% CI)3.77 [1.88, 7.58]

    4.4 Endocrine—treatment-emergent high prolactin at any time during treatment
1107Risk Ratio (M-H, Fixed, 95% CI)4.70 [2.25, 9.82]

    4.5 General deterioration—exacerbation of schizophrenia
1107Risk Ratio (M-H, Fixed, 95% CI)0.28 [0.09, 0.89]

    4.6 General deterioration—use of benzodiazepines
1107Risk Ratio (M-H, Fixed, 95% CI)0.57 [0.35, 0.92]

    4.7 Metabolic—increased appetite
1107Risk Ratio (M-H, Fixed, 95% CI)1.94 [0.59, 6.45]

    4.8 Metabolic—weight gain
1107Risk Ratio (M-H, Fixed, 95% CI)3.56 [1.14, 11.11]

    4.9 Metabolic—weight gain ≥ 5% of baseline
1202Risk Ratio (M-H, Fixed, 95% CI)4.41 [0.98, 19.91]

    4.10 Metabolic—weight gain ≥ 7% of baseline
1106Risk Ratio (M-H, Fixed, 95% CI)3.12 [1.34, 7.27]

    4.11 Metabolic—treatment-emergent high triglycerides at any time
1107Risk Ratio (M-H, Fixed, 95% CI)2.38 [1.31, 4.30]

 5 Adverse effects: 2. Different adverse effects (continuous measures—mean changes)2Mean Difference (IV, Fixed, 95% CI)Subtotals only

    5.1 Cardiovascular—corrected QT, QT/ms from baseline to end point (high score = poor
1107Mean Difference (IV, Fixed, 95% CI)-6.3 [-12.51, -0.09]

    5.2 Endocrine—prolactin, μg/L from baseline to end point (high score = poor)
2282Mean Difference (IV, Fixed, 95% CI)3.30 [-1.72, 8.31]

    5.3 Hepatic—ALT, U/L from baseline to end point (high score = poor)
1104Mean Difference (IV, Fixed, 95% CI)26.6 [11.34, 41.86]

    5.4 Hepatic—total bilirubin, mg/dL from baseline to end point (high score = poor)
1104Mean Difference (IV, Fixed, 95% CI)-3.70 [-6.30, -1.10]

    5.5 Metabolic—weight, kg from baseline to end point (high score = poor)
1106Mean Difference (IV, Fixed, 95% CI)4.2 [2.99, 5.41]

    5.6 Metabolic—BMI, kg/m² from baseline to end point (high score = poor)
1106Mean Difference (IV, Fixed, 95% CI)1.5 [1.06, 1.94]

    5.7 Metabolic—triglycerides, mg/dL from baseline to end point (high score = poor)
180Mean Difference (IV, Fixed, 95% CI)37.2 [8.88, 65.52]

    5.8 Renal—uric acid, μmol/L from baseline to end point (high score = poor)
1104Mean Difference (IV, Fixed, 95% CI)38.40 [18.88, 57.92]

 6 Leaving study early: 1. Various reasons (olanzapine vs placebo)1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    6.1 For any reason
1107Risk Ratio (M-H, Fixed, 95% CI)0.56 [0.36, 0.87]

    6.2 Because of adverse effects (elevated liver enzyme)
1107Risk Ratio (M-H, Fixed, 95% CI)5.42 [0.31, 95.43]

    6.3 Because of lack of efficacy
1107Risk Ratio (M-H, Fixed, 95% CI)0.27 [0.14, 0.52]

 7 Leaving study early: 2. Any reason (aripiprazole vs placebo)1202Risk Ratio (M-H, Fixed, 95% CI)1.76 [0.86, 3.63]

 8 Quality of life: 1. Mean end point PQ-LES-Q score at 6 weeks (data skewed, high score = good)Other dataNo numeric data

 
Analysis 1.3 Comparison 1 Atypical antipsychotics vs placebo (only short term), Outcome 3 Mental state: 2. Change in PANSS score (data skewed, high score = good).
Mental state: 2. Change in PANSS score (data skewed, high score = good)

StudyTreatmentMeanSDN

Findling 2008Aripiprazole 10mg6715799

Findling 2008Aripiprazole 30mg66.3164.997

Findling 2008Placebo73.8152.9998

 
Analysis 1.8 Comparison 1 Atypical antipsychotics vs placebo (only short term), Outcome 8 Quality of life: 1. Mean end point PQ-LES-Q score at 6 weeks (data skewed, high score = good).
Quality of life: 1. Mean end point PQ-LES-Q score at 6 weeks (data skewed, high score = good)

StudyInterventionMeanSDN

Findling 2008Aripiprazole 30mg50.29098

Findling 2008Placebo48.894.498

 
Comparison 2. Atypical vs typical antipsychotics (only short term)

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Global state: 1. Worse or no improvement121Risk Ratio (M-H, Fixed, 95% CI)3.3 [0.41, 26.81]

 2 Global state: 2a. Mean end point score (CGAS, high score = good)121Mean Difference (IV, Fixed, 95% CI)17.0 [7.74, 26.26]

 3 Global state: 2b. Mean end point score (CGI-I, high score = poor)1Mean Difference (IV, Random, 95% CI)Subtotals only

 4 Mental state: 1. No improvement (BPRS, high score = poor)2100Risk Ratio (M-H, Fixed, 95% CI)1.0 [0.38, 2.62]

 5 Mental state: 2a. Mean end point scores (various scales, high score = poor)5Mean Difference (IV, Fixed, 95% CI)Subtotals only

    5.1 B-HPRS
121Mean Difference (IV, Fixed, 95% CI)-3.60 [-6.64, -0.56]

    5.2 BPRS
5342Mean Difference (IV, Fixed, 95% CI)-1.34 [-3.24, 0.56]

    5.3 PANSS—total
1156Mean Difference (IV, Fixed, 95% CI)29.60 [20.84, 38.37]

    5.4 PANSS—positive subscale
1156Mean Difference (IV, Fixed, 95% CI)0.67 [-1.98, 3.32]

    5.5 PANSS—negative subscale
1156Mean Difference (IV, Fixed, 95% CI)1.83 [-1.55, 5.22]

 6 Mental state: 2b. Mean end point scores (high score = poor, skewed data)Other dataNo numeric data

    6.1 SAPS
Other dataNo numeric data

    6.2 SANS
Other dataNo numeric data

 7 Adverse effects: 1. Anticholinergic adverse effects (TESS)3Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    7.1 Total
140Risk Ratio (M-H, Fixed, 95% CI)0.2 [0.05, 0.80]

    7.2 Blood pressure—low
160Risk Ratio (M-H, Fixed, 95% CI)0.2 [0.01, 4.00]

    7.3 Dizziness
2100Risk Ratio (M-H, Fixed, 95% CI)0.27 [0.05, 1.60]

    7.4 Constipation
2100Risk Ratio (M-H, Fixed, 95% CI)0.14 [0.02, 1.12]

    7.5 Saliva—dry mouth
160Risk Ratio (M-H, Fixed, 95% CI)0.5 [0.10, 2.53]

    7.6 Saliva—hypersalivation
281Risk Ratio (M-H, Fixed, 95% CI)1.22 [0.49, 3.08]

    7.7 Tachycardia (resting heart rate ≥ 100 beats/min)
121Risk Ratio (M-H, Fixed, 95% CI)1.54 [0.72, 3.31]

 8 Adverse effects: 2a. Extrapyramidal adverse effects (TESS)2Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    8.1 Any
140Risk Ratio (M-H, Fixed, 95% CI)0.37 [0.20, 0.68]

    8.2 Restlessness
2100Risk Ratio (M-H, Fixed, 95% CI)0.71 [0.24, 2.10]

    8.3 Tremor
2100Risk Ratio (M-H, Fixed, 95% CI)0.46 [0.21, 1.04]

 9 Adverse effects: 2b. Extrapyramidal adverse effects (mean end point scores)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    9.1 AIMS (high score = poor)
121Mean Difference (IV, Fixed, 95% CI)-0.10 [-3.72, 3.52]

    9.2 S-ANRS (high score = poor)
121Mean Difference (IV, Fixed, 95% CI)-1.90 [-4.19, 0.39]

 10 Adverse effects: 3a. Other significant adverse effects2Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    10.1 Endocrine—mean end point serum prolactin concentration (mcg/L)
121Risk Ratio (M-H, Fixed, 95% CI)5.45 [0.29, 101.55]

    10.2 Haematology—drop in absolute neutrophil count below 1500 mm3
121Risk Ratio (M-H, Fixed, 95% CI)12.00 [0.75, 192.86]

    10.3 Central nervous system—somnolence/drowsiness (TESS)
281Risk Ratio (M-H, Fixed, 95% CI)1.19 [0.55, 2.55]

 11 Adverse effects: 3b. Other significant adverse effects (mean end point)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    11.1 Metabolic—body weight (kg)
1156Mean Difference (IV, Fixed, 95% CI)1.71 [-4.69, 8.11]

    11.2 Metabolic—serum cholesterol concentration (mg/dL)
1156Mean Difference (IV, Fixed, 95% CI)11.88 [-2.00, 25.76]

    11.3 Endocrine—serum prolactin concentration (mcg/L)
1156Mean Difference (IV, Fixed, 95% CI)1.71 [-4.69, 8.11]

 12 Leaving study early: various reasons4Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    12.1 For any reason
3187Risk Ratio (M-H, Fixed, 95% CI)0.62 [0.39, 0.97]

    12.2 Adverse effects—unspecified
3187Risk Ratio (M-H, Fixed, 95% CI)0.65 [0.36, 1.15]

    12.3 Adverse effects—drop in neutrophil count
121Risk Ratio (M-H, Fixed, 95% CI)5.45 [0.29, 101.55]

    12.4 Adverse effects—neuroleptic malignant syndrome
121Risk Ratio (M-H, Fixed, 95% CI)0.36 [0.02, 8.03]

    12.5 'Inadequate efficacy' or ≤ 20% reduction in CGI-I score
3421Risk Ratio (M-H, Fixed, 95% CI)0.59 [0.43, 0.82]

 
Analysis 2.6 Comparison 2 Atypical vs typical antipsychotics (only short term), Outcome 6 Mental state: 2b. Mean end point scores (high score = poor, skewed data).
Mental state: 2b. Mean end point scores (high score = poor, skewed data)

StudyTreatmentMeanSDN

SAPS

Kumra 1996Clozapine19.111.710

Kumra 1996Haloperidol35.915.611

SANS

Kumra 1996Clozapine46.030.310

Kumra 1996Haloperidol72.224.711

 
Comparison 3. Atypical antipsychotics vs atypical antipsychotics (only short term)

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Global state: 1. No improvement or no response4172Risk Ratio (M-H, Fixed, 95% CI)0.89 [0.65, 1.23]

    1.1 No improvement in CGI-S (risperidone vs olanzapine)
2111Risk Ratio (M-H, Fixed, 95% CI)1.04 [0.70, 1.54]

    1.2 No improvement in CGI (clozapine vs olanzapine)
139Risk Ratio (M-H, Fixed, 95% CI)0.5 [0.24, 1.03]

    1.3 No response (< 30% reduction in CGI-S score)
122Risk Ratio (M-H, Fixed, 95% CI)1.2 [0.52, 2.79]

 2 Global state: 2. Mean end point scores (high score = good)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    2.1 CGI-I
139Mean Difference (IV, Fixed, 95% CI)-1.07 [-1.92, -0.22]

    2.2 CGAS
139Mean Difference (IV, Fixed, 95% CI)4.10 [-6.71, 14.91]

 3 Mental state: 1. No response3Risk Difference (M-H, Fixed, 95% CI)Subtotals only

    3.1 Clozapine vs olanzapine (≤ 30% reduction in BPRS score)
139Risk Difference (M-H, Fixed, 95% CI)-0.44 [-0.72, -0.17]

    3.2 Olanzapine vs quetiapine (≤ 40% reduction in PANSS score)
120Risk Difference (M-H, Fixed, 95% CI)-0.20 [-0.62, 0.22]

    3.3 Quetiapine vs risperidone (≤ 30% reduction in PANSS score at 6 weeks)
122Risk Difference (M-H, Fixed, 95% CI)0.18 [-0.21, 0.58]

    3.4 Risperidone vs other atypical antipsychotics (≤ 40% improvement in PANSS total score)
129Risk Difference (M-H, Fixed, 95% CI)-0.27 [-0.64, 0.11]

    3.5 Risperidone vs olanzapine (≤ 40% reduction in PANSS score)
119Risk Difference (M-H, Fixed, 95% CI)-0.17 [-0.60, 0.27]

    3.6 Risperidone vs quetiapine (≤ 40% reduction in PANSS score)
119Risk Difference (M-H, Fixed, 95% CI)-0.37 [-0.79, 0.05]

 4 Mental state: mean end point scores (various scales, high score = poor)3Mean Difference (IV, Fixed, 95% CI)Totals not selected

    4.1 BPRS total
2Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.2 PANSS total
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.3 SANS total
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 5 Mental state: mean end point score (BPRS, data skewed, high score = poor)Other dataNo numeric data

 6 Adverse effects: 1a. Different adverse effects (clozapine vs olanzapine)1Odds Ratio (M-H, Fixed, 95% CI)Subtotals only

    6.1 Anticholinergic—salivation increase
139Odds Ratio (M-H, Fixed, 95% CI)6.0 [1.09, 33.02]

    6.2 Anticholinergic—sweating
139Odds Ratio (M-H, Fixed, 95% CI)9.5 [1.69, 53.33]

    6.3 Central nervous system—drowsiness
139Odds Ratio (M-H, Fixed, 95% CI)4.74 [0.21, 105.54]

    6.4 Metabolic—appetite increase
139Odds Ratio (M-H, Fixed, 95% CI)1.0 [0.26, 3.80]

    6.5 Metabolic—drug-induced diabetes
139Odds Ratio (M-H, Fixed, 95% CI)3.69 [0.14, 96.22]

    6.6 Metabolic—weight gain ≥ 7% of baseline body weight
139Odds Ratio (M-H, Fixed, 95% CI)1.90 [0.28, 12.87]

    6.7 Unspecified—"use of other antipsychotics"
128Odds Ratio (M-H, Fixed, 95% CI)0.46 [0.04, 5.77]

 7 Adverse effects: 1b. i. Different adverse effects (risperidone vs olanzapine)1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    7.1 Extrapyramidal—muscle stiffness/akathisia
119Risk Ratio (M-H, Fixed, 95% CI)2.22 [0.53, 9.37]

    7.2 Metabolic—weight gain ≥ 7% of baseline body weight
119Risk Ratio (M-H, Fixed, 95% CI)1.48 [0.85, 2.58]

 8 Adverse effects: 1b.ii. Different adverse effects—means at end of study (risperidone vs olanzapine, high score = poor)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    8.1 Endocrine—prolactin (mcg/L)
176Mean Difference (IV, Fixed, 95% CI)-2.30 [-9.97, 5.37]

 9 Adverse effects: 1b.iii. Different adverse effects—mean change (risperidone vs olanzapine, data skewed, high score = poor)Other dataNo numeric data

    9.1 Cardiac—QTc (ms)
Other dataNo numeric data

    9.2 Endocrine—prolactin (mcg/L)
Other dataNo numeric data

 10 Adverse effects: 1c. Different adverse effects (risperidone vs quetiapine)2Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    10.1 Endocrine—prolactin elevation
122Risk Ratio (M-H, Fixed, 95% CI)10.0 [1.53, 65.41]

    10.2 Extrapyramidal—akathisia—BARS
122Risk Ratio (M-H, Fixed, 95% CI)1.0 [0.40, 2.50]

    10.3 Extrapyramidal—general—AIMS
122Risk Ratio (M-H, Fixed, 95% CI)3.0 [0.37, 24.58]

    10.4 Extrapyramidal—general—SAS
122Risk Ratio (M-H, Fixed, 95% CI)2.0 [0.66, 6.04]

    10.5 Extrapyramidal—muscle stiffness/akathisia
119Risk Ratio (M-H, Fixed, 95% CI)4.44 [0.60, 32.77]

    10.6 Extrapyramidal—use of anticholinergic medication
122Risk Ratio (M-H, Fixed, 95% CI)5.0 [0.69, 36.13]

    10.7 Metabolic—weight gain ≥ 7% of baseline body weight
119Risk Ratio (M-H, Fixed, 95% CI)1.78 [0.92, 3.44]

 11 Adverse effects: 2a. Metabolic syndrome measures—means at end of study (clozapine vs olanzapine)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    11.1 BMI (kg/m2, high score = poor)
138Mean Difference (IV, Fixed, 95% CI)-0.5 [-3.35, 2.35]

    11.2 Cholesterol (mg/dL, high score = poor)
138Mean Difference (IV, Fixed, 95% CI)-10.40 [-31.59, 10.79]

    11.3 Glucose (mg/dL, high score = poor)
138Mean Difference (IV, Fixed, 95% CI)10.10 [1.46, 18.74]

    11.4 Triglycerides (mg/dL, high score = poor)
138Mean Difference (IV, Fixed, 95% CI)20.20 [-26.19, 66.59]

 12 Adverse effects: 2b.i. Metabolic syndrome measures—means at end of study (risperidone vs olanzapine)2Mean Difference (IV, Fixed, 95% CI)Subtotals only

    12.1 BMI (kg/m2, high score = poor)
134Mean Difference (IV, Fixed, 95% CI)-1.40 [-5.74, 2.94]

    12.2 Body weight (kg, high score = poor)
176Mean Difference (IV, Fixed, 95% CI)-2.30 [-9.97, 5.37]

    12.3 Cholesterol (mg/dL, high score = poor)
176Mean Difference (IV, Fixed, 95% CI)-27.10 [-50.13, -4.07]

 13 Adverse effects: 2b. ii. Metabolic syndrome measures—mean changes (risperidone vs olanzapine, data skewed, high score = poor)Other dataNo numeric data

    13.1 BMI (kg/m2)
Other dataNo numeric data

    13.2 Cholesterol (mg/dL)
Other dataNo numeric data

    13.3 Glucose (mg/dL)
Other dataNo numeric data

    13.4 Weight gain (kg)
Other dataNo numeric data

 14 Leaving study early: 1a. Clozapine vs olanzapine1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    14.1 For any reason
139Risk Ratio (M-H, Fixed, 95% CI)0.67 [0.23, 1.91]

    14.2 Because of neutropaenia (absolute neutrophil count = 1200)
139Risk Ratio (M-H, Fixed, 95% CI)0.39 [0.02, 8.93]

    14.3 Because of non-response
139Risk Ratio (M-H, Fixed, 95% CI)0.19 [0.03, 1.47]

    14.4 Because of weight gain
139Risk Ratio (M-H, Fixed, 95% CI)5.79 [0.30, 113.26]

 15 Leaving the study early: 1b. Olanzapine vs quetiapine2Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    15.1 For any reason
270Risk Ratio (M-H, Fixed, 95% CI)1.03 [0.55, 1.92]

    15.2 Because of non-response or lack of efficacy or worsening of clinical condition
270Risk Ratio (M-H, Fixed, 95% CI)0.47 [0.13, 1.74]

    15.3 Because of weight gain
120Risk Ratio (M-H, Fixed, 95% CI)3.0 [0.14, 65.90]

 16 Leaving study early: 1c. Risperidone vs olanzapine3Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    16.1 For any reason
3131Risk Ratio (M-H, Fixed, 95% CI)0.83 [0.52, 1.33]

    16.2 Because of non-response
3130Risk Ratio (M-H, Fixed, 95% CI)1.15 [0.44, 3.04]

    16.3 Because of weight gain
119Risk Ratio (M-H, Fixed, 95% CI)0.37 [0.02, 8.01]

    16.4 Because of adverse events
3130Risk Ratio (M-H, Fixed, 95% CI)1.21 [0.51, 2.87]

 17 Leaving the study early: 1d. Risperidone vs quetiapine1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    17.1 For any reason
120Risk Ratio (M-H, Fixed, 95% CI)0.25 [0.03, 1.86]

    17.2 Because of non-response
119Risk Ratio (M-H, Fixed, 95% CI)0.22 [0.01, 4.05]

    17.3 Because of weight gain
119Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 
Analysis 3.5 Comparison 3 Atypical antipsychotics vs atypical antipsychotics (only short term), Outcome 5 Mental state: mean end point score (BPRS, data skewed, high score = poor).
Mental state: mean end point score (BPRS, data skewed, high score = poor)

StudyInterventionMeanSDN

Sikich 2004Risperidone272019

Sikich 2004Olanzapine221216

Sikich 2008Risperidone29.623.241

Sikich 2008Olanzapine24.71635

 
Analysis 3.9 Comparison 3 Atypical antipsychotics vs atypical antipsychotics (only short term), Outcome 9 Adverse effects: 1b.iii. Different adverse effects—mean change (risperidone vs olanzapine, data skewed, high score = poor).
Adverse effects: 1b.iii. Different adverse effects—mean change (risperidone vs olanzapine, data skewed, high score = poor)

StudyInterventionMeanSDN

Cardiac—QTc (ms)

Sikich 2008Riesperidone0.529.541

Sikich 2008Olanzapine11.216.835

Endocrine—prolactin (mcg/L)

Sikich 2008Risperidone19.521.541

Sikich 2008Olanzapine-1.520.235

 
Analysis 3.13 Comparison 3 Atypical antipsychotics vs atypical antipsychotics (only short term), Outcome 13 Adverse effects: 2b. ii. Metabolic syndrome measures—mean changes (risperidone vs olanzapine, data skewed, high score = poor).
Adverse effects: 2b. ii. Metabolic syndrome measures—mean changes (risperidone vs olanzapine, data skewed, high score = poor)

StudyInterventionMeanSDN

BMI (kg/m2)

Sikich 2008Risperidone1.31.541

Sikich 2008Olanzapine2.21.235

Cholesterol (mg/dL)

Sikich 2008Risperidone-10.226.741

Sikich 2008Olanzapine19.923.935

Glucose (mg/dL)

Sikich 2008Risperidone1.27.341

Sikich 2008Olanzapine0.615.735

Weight gain (kg)

Sikich 2008Risperidone3.6441

Sikich 2008Olanzapine6.13.635

 
Comparison 4. Atypical (standard dose) vs atypical (low dose) antipsychotics (only short term)

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Global state: 1. No response—not sustained response at end of 8 weeks1255Risk Ratio (M-H, Fixed, 95% CI)0.59 [0.45, 0.77]

 2 Global state: 2a. Mean end point score (CGI-I, high score = poor)3468Mean Difference (IV, Fixed, 95% CI)-0.34 [-0.55, -0.13]

 3 Global state: 2b. Mean scores—at 6 weeks (aripiprazole 30 mg vs 10 mg)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    3.1 Change at CGAS score
1198Mean Difference (IV, Fixed, 95% CI)0.10 [-0.29, 0.49]

    3.2 Change in CGI-S score
1196Mean Difference (IV, Fixed, 95% CI)-0.10 [-0.13, -0.07]

    3.3 Mean end point CGI-I score
1196Mean Difference (IV, Fixed, 95% CI)-0.20 [-0.23, -0.17]

 4 Mental state: 1. No response2Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    4.1 ≤ 30% reduction in PANSS score
1255Risk Ratio (M-H, Fixed, 95% CI)0.78 [0.59, 1.03]

    4.2 ≤ 40% improvement in PANSS score
1255Risk Ratio (M-H, Fixed, 95% CI)0.82 [0.68, 0.98]

    4.3 Not achieving remission at 6 weeks on PANSS (aripiprazole 30 mg vs 10 mg)
1196Risk Ratio (M-H, Fixed, 95% CI)0.91 [0.66, 1.25]

 5 Mental state: 2. Mean end point scores (high score = poor)3Mean Difference (IV, Fixed, 95% CI)Subtotals only

    5.1 BPRS
117Mean Difference (IV, Fixed, 95% CI)-4.40 [-19.20, 10.40]

    5.2 PANSS
2451Mean Difference (IV, Fixed, 95% CI)-3.49 [-7.26, 0.28]

 6 Adverse effects: 1. Endrocrine1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    6.1 Prolactin—elevation above normal
1257Risk Ratio (M-H, Fixed, 95% CI)1.34 [1.04, 1.73]

    6.2 Prolactin—elevation > 100 ng/mL
1257Risk Ratio (M-H, Fixed, 95% CI)46.46 [6.50, 332.17]

    6.3 Prolactin—symptomatic hyperprolactinaemia
1257Risk Ratio (M-H, Fixed, 95% CI)3.70 [0.78, 17.45]

 7 Adverse effects: 2. Extrapyramidal2Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    7.1 Any
1254Risk Ratio (M-H, Fixed, 95% CI)3.31 [1.86, 5.87]

    7.2 Akathisia
2459Risk Ratio (M-H, Fixed, 95% CI)3.31 [1.46, 7.53]

    7.3 Dystonia
2459Risk Ratio (M-H, Fixed, 95% CI)2.17 [1.12, 4.18]

    7.4 Oculogyric crisis
1257Risk Ratio (M-H, Fixed, 95% CI)3.17 [0.13, 77.01]

    7.5 Parkinsonism
2459Risk Ratio (M-H, Fixed, 95% CI)2.32 [1.36, 3.98]

    7.6 Tremor
2459Risk Ratio (M-H, Fixed, 95% CI)4.27 [1.78, 10.24]

    7.7 Use of antiparkinsonian medications
1257Risk Ratio (M-H, Fixed, 95% CI)4.86 [1.91, 12.38]

 8 Adverse effects: 3. Metabolic1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    8.1 Weight gain—any
1257Risk Ratio (M-H, Fixed, 95% CI)1.24 [1.07, 1.44]

    8.2 Weight gain—moving from overweight to obese
142Risk Ratio (M-H, Fixed, 95% CI)1.96 [0.50, 7.67]

    8.3 Weight gain—"experiencing weight gain as an adverse effect"
1257Risk Ratio (M-H, Fixed, 95% CI)3.32 [1.47, 7.49]

 9 Adverse effects: 4. Others1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    9.1 Central nervous system—somnolence
1257Risk Ratio (M-H, Fixed, 95% CI)3.17 [1.68, 5.99]

    9.2 Non-specific—self-injury and aggression
1257Risk Ratio (M-H, Fixed, 95% CI)0.35 [0.04, 3.34]

    9.3 Non-specific—treatment-emergent adverse effects
1257Risk Ratio (M-H, Fixed, 95% CI)1.14 [0.97, 1.34]

    9.4 Non-specific—worsening of psychiatric symptoms
1257Risk Ratio (M-H, Fixed, 95% CI)1.32 [0.36, 4.80]

 10 Adverse effects: 5. Others (particular to the aripiprazole 30 mg vs 10 mg comparison)1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    10.1 Central nervous system—somnolence
1202Risk Ratio (M-H, Fixed, 95% CI)1.96 [1.00, 3.83]

    10.2 Extrapyramidal—any adverse effects
1202Risk Ratio (M-H, Fixed, 95% CI)1.66 [0.89, 3.11]

    10.3 Extrapyramidal—akathisia
1202Risk Ratio (M-H, Fixed, 95% CI)1.96 [0.77, 5.02]

    10.4 Extrapyramidal—dystonia
1202Risk Ratio (M-H, Fixed, 95% CI)0.49 [0.09, 2.62]

    10.5 Extrapyramidal—parkinsonism (SAS)
1202Risk Ratio (M-H, Fixed, 95% CI)2.03 [1.17, 3.52]

    10.6 Extrapyramidal—tremor
1202Risk Ratio (M-H, Fixed, 95% CI)5.88 [1.35, 25.62]

    10.7 Metabolic—weight gain of ≥ 5% of body weight
1196Risk Ratio (M-H, Fixed, 95% CI)0.84 [0.36, 1.93]

 11 Adverse effects: 6. Laboratory tests (particular to the aripiprazole 30 mg vs 10 mg comparison)Other dataNo numeric data

    11.1 Change in serum cholesterol concentration (data skewed, high score = poor)
Other dataNo numeric data

    11.2 Change in serum prolactin concentration (data skewed, high score = poor)
Other dataNo numeric data

 12 Leaving study early: 1. Various reasons1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    12.1 For any reason
1257Risk Ratio (M-H, Fixed, 95% CI)0.74 [0.52, 1.06]

    12.2 Adverse effects
1257Risk Ratio (M-H, Fixed, 95% CI)0.88 [0.28, 2.81]

    12.3 Insufficient response
1257Risk Ratio (M-H, Fixed, 95% CI)0.77 [0.45, 1.32]

    12.4 Other reasons
1257Risk Ratio (M-H, Fixed, 95% CI)0.79 [0.18, 3.47]

 13 Leaving study early: 2. Ziprasidone 160 mg vs 80 mg117Risk Ratio (M-H, Fixed, 95% CI)2.22 [0.58, 8.44]

 14 Leaving study early: 3. Aripiprazole 30 mg vs aripiprazole 10 mg1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    14.1 For any reason
1202Risk Ratio (M-H, Fixed, 95% CI)1.10 [0.60, 2.04]

    14.2 Adverse effects
1202Risk Ratio (M-H, Fixed, 95% CI)0.56 [0.17, 1.85]

    14.3 Lack of efficacy
1202Risk Ratio (M-H, Fixed, 95% CI)0.20 [0.02, 1.65]

 15 Quality of life: mean end point score—at 6 weeks (PQ-LES-Q, aripiprazole 30 mg vs aripiprazole 10 mg, data skewed, high score = good)Other dataNo numeric data

 
Analysis 4.11 Comparison 4 Atypical (standard dose) vs atypical (low dose) antipsychotics (only short term), Outcome 11 Adverse effects: 6. Laboratory tests (particular to the aripiprazole 30 mg vs 10 mg comparison).
Adverse effects: 6. Laboratory tests (particular to the aripiprazole 30 mg vs 10 mg comparison)

StudyInterventionMeanSDN

Change in serum cholesterol concentration (data skewed, high score = poor)

Findling 2008Aripiprazole 30mg-5.0123.2895

Findling 2008Aripiprazole 10mg-7.4327.9998

Change in serum prolactin concentration (data skewed, high score = poor)

Findling 2008Aripiprazole 30mg-15.1426.8792

Findling 2008Aripiprazole 10mg-11.9323.2998

 
Analysis 4.15 Comparison 4 Atypical (standard dose) vs atypical (low dose) antipsychotics (only short term), Outcome 15 Quality of life: mean end point score—at 6 weeks (PQ-LES-Q, aripiprazole 30 mg vs aripiprazole 10 mg, data skewed, high score = good).
Quality of life: mean end point score—at 6 weeks (PQ-LES-Q, aripiprazole 30 mg vs aripiprazole 10 mg, data skewed, high score = good)

StudyInterventionMeanSDN

Findling 2008Aripiprazole 30mg50.29097

Findling 2008Aripiprazole 10mg50.190.999

 
Summary of findings for the main comparison. Atypical antipsychotics compared with placebo (only short term)

Atypical antipsychotics compared with placebo (only short term)

Patient or population: individuals with psychosis
Settings:
Intervention: atypical antipsychotics
Comparison: placebo

OutcomesIllustrative comparative risks* (95% CI)Risk ratio
(95% CI)
No of participants
(studies)
Quality of the evidence
(GRADE)
Comments

Assumed riskCorresponding risk

PlaceboAtypical antipsychotics

Weight gainStudy populationRR 3.56
(1.14 to 11.11)
107
(1 study)
⊕⊕⊕⊝
moderatea

86 per 1000305 per 1000
(98 to 952)

Moderate

86 per 1000306 per 1000
(98 to 955)

Weight gain7% of baselineStudy populationRR 3.12
(1.34 to 7.27)
106
(1 study)
⊕⊕⊕⊕
high

147 per 1000459 per 1000
(197 to 1000)

Moderate

147 per 1000459 per 1000
(197 to 1000)

High prolactin at any time during treatmentStudy populationRR 4.7
(2.25 to 9.82)
107
(1 studies)
⊕⊕⊕⊕
high

171 per 1000806 per 1000
(386 to 1000)

Moderate

171 per 1000804 per 1000
(385 to 1000)

Change in corrected QTMean change in corrected QT in the intervention groups was 6.3 lower (12.51 to 0.09 lower)107
(1 study)
⊕⊕⊕⊕
high

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the risk ratio of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence.
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

 aWide confidence interval.
 
Summary of findings 2. Atypical compared with typical antipsychotics (only short term)

Atypical compared with typical antipsychotics (only short term)

Patient or population: individuals with psychosis
Settings:
Intervention: atypical antipsychotics
Comparison: typical antipsychotics

OutcomesIllustrative comparative risks* (95% CI)Risk ratio
(95% CI)
No of participants
(studies)
Quality of the evidence
(GRADE)
Comments

Assumed riskCorresponding risk

Typical antipsychoticsAtypical antipsychotics

Worse or no improvementStudy populationRR 3.3
(0.41 to 26.81)
21
(1 study)
⊕⊕⊝⊝
lowa,b

91 per 1000300 per 1000
(37 to 1000)

Moderate

91 per 1000300 per 1000
(37 to 1000)

Anticholinergic adverse effectsStudy populationRR 0.2
(0.05 to 0.8)
40
(1 study)
⊕⊕⊕⊝
moderatec

500 per 1000100 per 1000
(25 to 400)

Moderate

500 per 1000100 per 1000
(25 to 400)

Drop in the absolute neutrophil count below 1500 per mm3 Study populationRR 12
(0.75 to 192.86)
21
(1 study)
⊕⊝⊝⊝
very lowb,d

0 per 10000 per 1000
(0 to 0)

Moderate

0 per 10000 per 1000
(0 to 0)

Leaving the study early because of adverse effectsStudy populationRR 3.3
(0.41 to 26.81)
21
(1 study)
⊕⊕⊕⊝
moderatea

91 per 1000300 per 1000
(37 to 1000)

Moderate

91 per 1000300 per 1000
(37 to 1000)

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the risk ratio of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence.
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

 aWide confidence interval.
bThe high incidence of neutropenia in the clozapine group has been reported but not adequately discussed.
cAlthough one of the studies said it was randomised and double blind, the authors did not provide a description.
dVery wide confidence interval.
 
Summary of findings 3. Atypical vs atypical antipsychotics (only short term)

Atypical vs atypical antipsychotics (only short term)

Patient or population: individuals with psychosis
Settings:
Intervention: Atypical antipsychotics
Comparison: Atypical antipsychotics

OutcomesIllustrative comparative risks* (95% CI)Risk ratio
(95% CI)
No of participants
(studies)
Quality of the evidence
(GRADE)
Comments

Assumed riskCorresponding risk

Atypical antipsychoticsAtypical antipsychotics

No improvement (risperidone vs olanzapine)Study populationRR 0.50
(0.24 to 1.07)
111
(2 studies)
⊕⊕⊕⊝
moderatea

451 per 1000468 per 1000
(291 to 654)

Moderate

413 per 1000429 per 1000
(260 to 618)

No improvement (clozapine vs olanzapine)Study populationRR 0.5
(0.24 to 1.03)
39
(1 study)
⊕⊕⊕⊕
high

667 per 1000333 per 1000
(160 to 687)

Moderate

667 per 1000333 per 1000
(160 to 687)

Use of other antipsychoticsStudy populationRR 0.5
(0.05 to 4.9)
28
(1 study)
⊕⊕⊕⊝
moderateb

143 per 100071 per 1000
(7 to 700)

Moderate

143 per 100072 per 1000
(7 to 701)

Drug-induced diabetes (clozapine vs olanzapine)Study populationRR 3
(0.13 to 67.91)
28
(1 study)
⊕⊕⊕⊝
moderateb

0 per 10000 per 1000
(0 to 0)

Moderate

0 per 10000 per 1000
(0 to 0)

Elevated prolactin (risperidone vs quetiapine)Study populationRR 10
(1.53 to 65.41)
22
(1 study)
⊕⊕⊕⊝
moderatec

91 per 1000909 per 1000
(139 to 1000)

Moderate

91 per 1000910 per 1000
(139 to 1000)

Weight gain in kg (risperidone vs olanzapine)The mean weight gain in kg (risperidone vs olanzapine) in the intervention groups was
2.5 lower
(4.21 to 0.79 lower)
76
(1 study)
⊕⊕⊕⊕
high

Leaving the study early because of weight gain (risperidone vs olanzapine)Study populationRR 0.37
(0.02 to 8.01)
19
(1 study)
⊕⊕⊝⊝
lowb,d

100 per 100037 per 1000
(2 to 801)

Moderate

100 per 100037 per 1000
(2 to 801)

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the risk ratio of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence.
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

 aRisk ratio reduction (RRR) or risk ratio increase (RRI) greater than 25%.
bWide confidence interval.
cOpen-label study with blind midpoint and end point assessments.
dThis was an open-label study.
 
Summary of findings 4. Atypical (standard-dose) vs atypical (low-dose) antipsychotics (only short term)

Atypical (standard-dose) vs atypical (low-dose) antipsychotics (only short term)

Patient or population: individuals with psychosis
Settings:
Intervention: atypical antipsychotics (standard dose)
Comparison: atypical antipsychotics (low dose)

OutcomesIllustrative comparative risks* (95% CI)Risk ratio
(95% CI)
No of participants
(studies)
Quality of the evidence
(GRADE)
Comments

Assumed riskCorresponding risk

Atypical antipsychotics (low dose)Atypical antipsychotics (standard dose)

No responseStudy populationRR 0.54
(0.38 to 0.75)
255
(1 study)
⊕⊕⊕⊕
high

496 per 1000268 per 1000
(189 to 372)

Moderate

496 per 1000268 per 1000
(188 to 372)

Symptomatic hyperprolactinaemiaSee commentSee commentNot estimable257
(1 study)
⊕⊕⊕⊕
high

Use of antiparkinsonian medicationsStudy populationRR 4.86
(1.91 to 12.38)
257
(1 study)
⊕⊕⊕⊕
high

38 per 1000184 per 1000
(72 to 469)

Moderate

38 per 1000185 per 1000
(73 to 470)

Weight gain (standard-dose vs low-dose risperidone)Study populationRR 3.32
(1.47 to 7.49)
257
(1 study)
⊕⊕⊕⊕
high

53 per 1000176 per 1000
(78 to 397)

Moderate

53 per 1000176 per 1000
(78 to 397)

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence.
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.