Quetiapine versus typical antipsychotic medications for schizophrenia

  • Review
  • Intervention

Authors


Abstract

Background

Quetiapine is a widely used atypical antipsychotic drug for schizophrenia that has been on the market for over a decade. However, It is not clear how the effects of quetiapine differ from typical antipsychotics.

Objectives

To review the effects of quetiapine in comparison with typical antipsychotics in the treatment of schizophrenia and schizophrenia-like psychosis.

Search methods

We searched the Cochrane Schizophrenia Group Trials Register (March 2010), and inspected references of all identified studies.

Selection criteria

We included all randomised control trials comparing oral quetiapine with typical antipsychotic drugs in people with schizophrenia or schizophrenia-like psychosis.

Data collection and analysis

We extracted data independently. For dichotomous data, we calculated risk ratio (RR) and 95% confidence intervals (CI) using a random-effects model. We presented chosen outcomes in a 'Summary of findings' table and comparative risks where appropriate. For continuous data, we calculated mean differences (MD) based on a random-effects model. We assessed risk of bias for included studies.

Main results

The review includes 43 randomised controlled trials (RCTs) with 7217 participants. Most studies were from China. The percentages of participants leaving the studies early were similar (36.5% in quetiapine group and 36.9% in typical antipsychotics group) and no significant difference between groups was apparent for leaving early due to any reason (23 RCTs n = 3576 RR 0.91 CI 0.81 to 1.01, moderate quality evidence), however, fewer participants in the quetiapine group left the studies early due to adverse events (15 RCTs, n = 3010, RR 0.48 CI 0.30 to 0.77).

Overall global state was similar between groups (no clinically significant response; 16 RCTs, n = 1607, RR 0.96 CI 0.75 to 1.23, moderate quality evidence) and there was no significant difference in positive symptoms (PANSS positive subscore: 22 RCTs, n = 1934, MD 0.02 CI -0.39 to 0.43, moderate quality evidence). General psychopathology was equivocal (PANSS general psychopathology subscore: 18 RCTs, n = 1569, MD -0.20 CI -0.83 to 0.42) between those allocated to quetiapine and typical antipsychotics. However, quetiapine was statistically significantly more efficacious for negative symptoms (PANSS negative subscore: 22 RCTs, n = 1934, MD -0.82 CI -1.59 to -0.04, moderate quality evidence), however, this result was highly heterogeneous and driven by two small outlier studies with high effect sizes. Without these two studies, there was no heterogeneity and no statistically significant difference between quetiapine and typical antipsychotics.

Compared with typical antipsychotics, quetiapine might cause fewer adverse effects (9 RCTs, n = 1985, RR 0.76 CI 0.64 to 0.90 number needed to treat to induce harm (NNTH) 10, CI 8 to 17), less abnormal ECG (2 RCTs, n = 165, RR 0.38 CI 0.16 to 0.92, NNTH 8, CI 4 to 55), fewer overall extrapyramidal effects (8 RCTs, n = 1,095, RR 0.17 CI 0.09 to 0.32, NNTH 3, CI 3 to 3, moderate quality evidence) and fewer specific extrapyramidal effects including akathisia, parkinsonism, dystonia and tremor. Moreover, it might cause lower prolactin level (4 RCTs, n = 1034, MD -16.20 CI -23.34 to -9.07, moderate quality evidence) and less weight gain compared with some typical antipsychotics in the short term (9 RCTs, n = 866, RR 0.52 CI 0.34 to 0.80, NNTH 8, CI 6 to 15).

However, there was no significant difference between the two groups in suicide attempt, suicide, death, QTc prolongation, low blood pressure, tachycardia, sedation, gynaecomastia, galactorrhoea, menstrual irregularity and white blood cell count.

Authors' conclusions

Quetiapine may not differ from typical antipsychotics in the treatment of positive symptoms and general psychopathology. There are no clear differences in terms of the treatment of negative symptoms. However, it causes fewer adverse effects in terms of abnormal ECG, extrapyramidal effects, abnormal prolactin levels and weight gain.

Résumé scientifique

La quétiapine versus les traitements médicamenteux antipsychotiques typiques de la schizophrénie

Contexte

La quétiapine est un antipsychotique atypique largement utilisé pour la schizophrénie et qui est sur le marché depuis plus d'une décennie. Il n'est cependant pas clair en quoi les effets de la quétiapine diffèrent de ceux des antipsychotiques typiques.

Objectifs

Comparer les effets de la quétiapine et des antipsychotiques typiques dans le traitement de la schizophrénie et de la psychose de type schizophrénique.

Stratégie de recherche documentaire

Nous avons effectué une recherche dans le registre d'essais cliniques du groupe Cochrane sur la schizophrénie (mars 2010) et nous avons examiné les références bibliographiques de toutes les études identifiées.

Critères de sélection

Nous avons inclus tout essai contrôlé randomisé comparant la quétiapine orale à des médicaments antipsychotiques typiques chez des personnes atteintes de schizophrénie ou de psychose de type schizophrénique.

Recueil et analyse des données

Nous avons extrait les données de manière indépendante. Pour les données dichotomiques, nous avons calculé le risque relatif (RR) et les intervalles de confiance (IC) à 95 % à l'aide d'un modèle à effets aléatoires. Nous avons présenté des critères de résultat choisis dans une table « Résumé des conclusions » et, le cas échéant, les risques comparatifs. Pour les variables continues, nous avons calculé la différence moyenne (DM), sur la base d'un modèle à effets aléatoires. Nous avons évalué le risque de biais des études incluses.

Résultats principaux

Cette revue inclut 43 essais contrôlés randomisés (ECR) totalisant 7217 participants. La plupart des études provenaient de Chine. Les pourcentages de participants ayant abandonné les études avant terme étaient similaires (36,5 % dans le groupe de la quétiapine et 36,9 % dans celui des antipsychotiques typiques) et aucune différence significative n'est apparue entre les groupes concernant l'abandon avant terme pour une raison spécifique quelconque (23 ECR, n = 3576 ; RR 0,91 [IC 0,81 à 1,01] ; données de qualité moyenne). Toutefois, moins de participants dans le groupe à quétiapine avaient abandonné les études avant terme pour cause d'événements indésirables (15 ECR, n = 3010 ; RR 0,48 [IC 0,30 à 0,77]).

L'état global était dans l'ensemble similaire entre les groupes (pas de réponse cliniquement significative ; 16 ECR, n = 1607 ; RR 0,96 [IC 0,75 à 1,23] ; données de qualité modérée) et il n'y avait pas de différence significative quant aux symptômes positifs (sous-score positif PANSS: 22 ECR, n = 1934 ; DM 0,02 [IC -0,39 à 0,43] ; données de qualité modérée). La psychopathologie générale était équivoque (sous-score PANSS de psychopathologie générale : 18 ECR, n = 1569 ; DM -0,20 [IC -0,83 à 0,42]) entre ceux assignés à la quétiapine et ceux assignés aux antipsychotiques typiques. La quétiapine était cependant plus efficace de manière statistiquement significative pour les symptômes négatifs (sous-score PANSS négatif : 22 ECR, n = 1934 ; DM -0,82 [IC -1,59 à -0,04] ; données de qualité modérée), mais ce résultat était très hétérogène et causé par deux petites études aberrantes dans lesquelles les effets étaient de grande ampleur. Sans ces deux études, il n'y avait ni hétérogénéité ni différence statistiquement significative entre la quétiapine et les antipsychotiques typiques.

En comparaison avec les antipsychotiques typiques, la quétiapine semble causer moins d'effets indésirables (9 ECR, n = 1985 ; RR 0,76 [IC 0,64 à 0,90] ; nombre nécessaire à traiter pour nuire (NNN) 10 [IC 8 à 17]), moins d'ECG anormaux (2 ECR, n = 165 ; RR 0,38 [IC 0,16 à 0,92] ; NNN 8 [IC 4 à 55]), moins d'effets extrapyramidaux globaux (8 ECR, n = 1 095 ; RR 0,17 [IC 0,09 à 0,32] ; NNN 3 [IC 3 à 3], données de qualité modérée) et moins d'effets extrapyramidaux spécifiques, comme l'akathisie, le parkinsonisme, la dystonie et la trépidation. Il se pourrait en outre qu'elle cause à court terme un taux moins élevé de prolactine (4 ECR, n = 1034 ; DM -16,20 [IC -23,34 à -9,07] ; données de qualité modérée) et une moindre prise de poids que certains antipsychotiques typiques (9 ECR, n = 866 ; RR 0,52 [0,34 à 0,80] ; NNN 8 [IC 6 à 15]).

Il n'y avait toutefois pas de différence significative entre les deux groupes pour les tentatives de suicide, le suicide, le décès, l'allongement du QTc, l'hypotension artérielle, la tachycardie, la sédation, la gynécomastie, la galactorrhée, les irrégularités menstruelles et le nombre de globules blancs.

Conclusions des auteurs

La quétiapine ne semble pas différer des antipsychotiques typiques pour le traitement des symptômes positifs et la psychopathologie générale. Il n'y a pas de nettes différences pour ce qui est du traitement des symptômes négatifs. Elle provoque cependant moins d'effets indésirables en termes d'ECG anormal, d'effets extrapyramidaux, de taux de prolactine anormaux et de prise de poids.

Plain language summary

Quetiapine versus typical antipsychotic drugs for schizophrenia

Antipsychotic drugs are the main treatment for schizophrenia, helping to treat both the positive symptoms (such as hearing voices, seeing things and having strange beliefs) and negative symptoms (including apathy, tiredness and loss of emotion) of this illness. Selecting the most effective antipsychotic drug that can be tolerated by people with schizophrenia is crucial to successful treatment. Older drugs (also known as typical or first generation antipsychotic drugs), such as chlorpromazine and haloperidol, have been used in treating schizophrenia for over 50 years. Although these older drugs are good at treating the positive symptoms of schizophrenia they tend to cause undesirable side effects. These side effects can mean that people do not tolerate or like taking these drugs, which may lead to relapse and admission to hospital. Since 1988, a newer generation of antipsychotic drugs has become available. These new drugs (known as atypical or second generation antipsychotic drugs) are effective in treating the symptoms of schizophrenia but thought to have less side effects than older drugs. However, although newer drugs may cause less side effects such as movement disorders, they have been linked to other side effects like heart problems or weight gain. Quetiapine is a new antipsychotic drug for schizophrenia that has been available for over a decade. However, it is not clear how the effects of quetiapine differ from older antipsychotic drugs. This review evaluated the effectiveness and tolerability of quetiapine versus older antipsychotic drugs. The review included 43 trials with a total of 7217 people. Most studies were from China. In the main, quetiapine did not differ from older drugs for the treatment of positive symptoms of mental illness. There were also no clear differences in terms of the treatment of negative symptoms. However, it is important to note that evidence from these trials suggests quetiapine causes fewer side effects (such as weight gain, dizziness, movement disorders, the inability to sit still, shaking, tremors and abnormal levels of the hormone prolactin, which can contribute to sexual and mental health problems). However, evidence from the trials is limited due to high numbers of people leaving early in almost all of the studies. More evidence through the completion of well designed studies comparing quetiapine with older antipsychotic drugs is needed.

This plain language summary has been written by a consumer, Benjamin Gray, Service User: RETHINK.

Résumé simplifié

La quétiapine versus les médicaments antipsychotiques typiques pour la schizophrénie

Les médicaments antipsychotiques constituent le principal traitement de la schizophrénie, agissant à la fois sur les symptômes positifs (comme entendre des voix, voir des choses et croire des choses étranges) et sur les symptômes négatifs (notamment l'apathie, la fatigue et la perte d'émotion) de cette maladie. Le choix du médicament antipsychotique le plus efficace pouvant être toléré par la personne souffrant de schizophrénie est crucial pour la réussite du traitement. Les médicaments les plus anciens (également appelés antipsychotiques typiques ou de première génération), tels que la chlorpromazine et l'halopéridol, sont utilisés depuis 50 ans dans le traitement de la schizophrénie. Bien que ces médicaments anciens soient efficaces pour traiter les symptômes positifs de la schizophrénie, ils ont tendance à provoquer des effets secondaires indésirables. Ces effets secondaires peuvent faire que les personnes ne tolèrent pas ou n'aiment pas prendre ces médicaments, ce qui peut conduire à la rechute et à l'hospitalisation. Depuis 1988, on dispose d'une nouvelle génération de médicaments antipsychotiques. Ces nouveaux médicaments (appelés antipsychotiques atypiques ou de deuxième génération) sont efficaces pour traiter les symptômes de la schizophrénie mais considérés avoir moins d'effets secondaires que les médicaments plus anciens. Cependant, bien que les nouveaux médicaments provoquent moins d'effets secondaires (moins de troubles du mouvement par exemple), on les a associés à d'autres effets secondaires tels que des problèmes cardiaques ou de prise de poids. La quétiapine est un nouveau médicament antipsychotique pour la schizophrénie qui est disponible depuis plus d'une décennie. Il n'est cependant pas clair en quoi la quétiapine diffère dans ses effets des antipsychotiques plus anciens. Cette revue a évalué l'efficacité et la tolérabilité de la quétiapine en comparaison avec les anciens antipsychotiques. La revue a inclus 43 essais totalisant 7217 personnes. La plupart des études provenaient de Chine. Pour l'essentiel, la quétiapine ne différait pas des anciens médicaments pour ce qui est du traitement des symptômes positifs de la maladie mentale. Il n'y avait pas non plus de différences claires pour ce qui est du traitement des symptômes négatifs. Cependant, il est important de noter que les résultats de ces essais suggèrent que la quétiapine provoque moins d'effets secondaires (tels que gain de poids, vertiges, troubles du mouvement, incapacité à rester assis, tremblements, trépidation et niveaux anormaux de l'hormone prolactine, qui peuvent contribuer à des problèmes de santé sexuelle et mentale). Toutefois, les données fournies par les essais sont limitées en raison du nombre élevé de personnes ayant abandonné avant terme dans presque toutes les études. On a besoin d'études bien conçues comparant la quétiapine à des médicaments antipsychotiques plus anciens, afin d'obtenir des données supplémentaires

Ce résumé en langage simplifié a été rédigé par un consommateur, Benjamin Gray, Bénéficiaire du service : RETHINK.

Notes de traduction

Traduit par: French Cochrane Centre 3rd June, 2013
Traduction financée par: Pour la France : Minist�re de la Sant�. Pour le Canada : Instituts de recherche en sant� du Canada, minist�re de la Sant� du Qu�bec, Fonds de recherche de Qu�bec-Sant� et Institut national d'excellence en sant� et en services sociaux.

Laički sažetak

Usporedba kvetiapina i tipičnih antipsihotika za shizofreniju

Antipsihotici su glavna terapija shizofrenije te pomažu u liječenju i pozitivnih (poput zvučnih i vizualnih halucinacija i neobičnih uvjerenja) i negativnih simptoma (uključujući apatiju, umor i gubitak emocija) te bolesti. Za uspješno liječenje ključan je izbor najučinkovitijeg antipsihotika koji će pacijenti dobro podnositi.Stariji lijekovi (poznati i kao tipični ili antipsihotici prve generacije), poput kloropromazina i haloperidola, koriste se u liječenju shizofrenije već preko 50 godina. Iako su ti lijekovi učinkoviti u liječenju pozitivnih simptoma shizofrenije, skloni su izazivanju nepoželjnih nuspojava.Te nuspojave mogu značiti da osobe ne podnose ili ne vole uzimati te lijekove, što može dovesti do ponovnog javljanja simptoma (relapsa) i prijema u bolnicu. Od 1988. dostupna je nova generacija antipsihotika. Ti novi lijekovi (poznati i kao atipični ili antipsihotici druge generacije) su učinkoviti u liječenju simptoma shizofrenije, ali se smatra da imaju manje nuspojava od starijih lijekova.Međutim, iako noviji lijekovi uzrokuju manje nuspojava poput poremećaja kretanja, povezivani su s drugim nuspojavama poput srčanih problema i povećanjem tjelesne mase. Kvetiapin je novi antipsihotik koji je dostupan više od desetljeća. Međutim, nije jasno kako se učinci kvetiapina razlikuju od učinaka starijih antipsihotika. Ovaj Cochrane sustavni pregled procijenio je učinkovitost i podnošljivost kvetiapina u usporedbi sa starijim antipsihoticima. Pregled je uključio 43 studije s ukupno 7217 sudionika. Većina studija je provedena u Kini. Uglavnom, kvetiapin se nije razlikovao od starijih lijekova u terapiji pozitivnih simptoma mentalne bolesti. Također, nije bilo jasnih razlika u pogledu terapije negativnih simptoma. Međutim, važno je spomenuti da dokazi iz tih studija upućuju da kvetiapin uzrokuje manje nuspojava (poput povećanja tjelesne mase, vrtoglavice, poremećaja kretanja, nesposobnosti mirnog sjedenja, drhtanja, tremora i abnormalnih količina hormona prolaktina, koji može doprinijeti nastanku seksualnih i problema s mentalnim zdravljem). Međutim, dokazi iz ih studija su ograničeni zbog velikog broja osoba koje su studiju napustile rano u gotovo svim studijama. Potrebno je više dokaza kroz završetak dobro dizajniranih studija koje uspoređuju kvetiapin sa starijim antipsihoticima.

Ovaj sažetak na jednostavnom jeziku je napisao Benjamin Gray predstavnik udruge pacijenata Rethink Mental Illness

Bilješke prijevoda

Hrvatski Cochrane
Preveo: Adam Galkovski
Ovaj sažetak preveden je u okviru volonterskog projekta prevođenja Cochrane sažetaka. Uključite se u projekt i pomozite nam u prevođenju brojnih preostalih Cochrane sažetaka koji su još uvijek dostupni samo na engleskom jeziku. Kontakt: cochrane_croatia@mefst.hr

Summary of findings(Explanation)

Summary of findings for the main comparison. Quetiapine compared to Typical antipsychotics for schizophrenia
  1. 1 Limitations in design - rated 'serious' : poor description of randomisation, no details about blinding, no details about concealment.
    2 Inconsistency -rated 'serious' : the measurement of cognitive function were various.
    3 Imprecision - rated 'serious' : number of participants was very small.

    PANSS: Positive and Negative Syndrome Scale

Quetiapine compared to Typical antipsychotics for schizophrenia
Patient or population: patients with schizophrenia
Settings:
Intervention: Quetiapine
Comparison: Typical antipsychotics
OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments
Assumed riskCorresponding risk
Typical antipsychoticsQuetiapine
Global state
No clinical significant response as defined by the individual studies
Study populationRR 0.96
(0.75 to 1.23)
1607
(16 studies)
⊕⊕⊕⊝
moderate 1
 
145 per 1000139 per 1000
(109 to 178)
Moderate
  
Leaving the study early due to any reasonStudy populationRR 0.91
(0.81 to 1.01)
3576
(23 studies)
⊕⊕⊕⊝
moderate 1
 
369 per 1000336 per 1000
(299 to 373)
Moderate
  
Positive symptoms
PANSS positive subscore
 The mean positive symptoms in the intervention groups was
0.02 higher
(0.39 lower to 0.43 higher)
 1934
(22 studies)
⊕⊕⊕⊝
moderate 1
 
Negative symptoms
PANSS negative subscore
 The mean negative symptoms in the intervention groups was
0.82 lower
(1.59 to 0.04 lower)
 1934
(22 studies)
⊕⊕⊕⊝
moderate 1
 
Cognitive function
Average endpoint scores as defined by the original studies
 The mean cognitive function in the intervention groups was
1.55 higher
(0.62 lower to 3.72 higher)
 142
(3 studies)
⊕⊝⊝⊝
very low 1,2,3
 
Extrapyramidal effectsStudy populationRR 0.17
(0.09 to 0.32)
1095
(8 studies)
⊕⊕⊕⊝
moderate 1
 
548 per 100093 per 1000
(49 to 175)
Moderate
  
Prolactin level
Average level in ng/mL
 The mean prolactin level in the intervention groups was
16.20 lower
(23.34 to 9.07 lower)
 1034
(4 studies)
⊕⊕⊕⊝
moderate 1
 
*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.

Background

Description of the condition

Schizophrenia is often a chronic and disabling psychiatric disorder. It affects approximately one per cent of the population. The severity of symptoms often causes substantial and long-lasting impairments. Moreover, the financial cost of treatment in schizophrenia is high for both afflicted people and health services as it often requires hospitalisation (Buchanan 2005).

Description of the intervention

Antipsychotics have been the core treatment for schizophrenia, thus selecting the most effective and tolerable antipsychotic is key to maximizing treatment outcomes. Typical antipsychotics such as chlorpromazine and haloperidol, although they have been used in treating schizophrenia for over 50 years, tend to cause undesirable adverse effects that may lead to non-compliance. Quetiapine is a widely used atypical antipsychotic drug for schizophrenia which has been on the market for over a decade. However, It is not clear how the effects of quetiapine differ from typical antipsychotics.

How the intervention might work

Experimental laboratory studies have indicated that quetiapine is a clozapine-like atypical antipsychotic (Goldstein 1993; Migler 1993; Saller 1993). While olanzapine, risperidone, sertindole and ziprasidone have high affinities (< 50 nM) to both D2 and 5-HT2A receptors, quetiapine is similar to clozapine in having only moderate affinities (< 500 nM) to these sites (Goldstein 1995) but a high affinity to histamine receptors (< 50 nM) (Srisurapanont 2004). As an agent with moderate affinities to dopamine D2 and serotonin 5-HT2A receptors, quetiapine is less likely to cause extrapyramidal side effects and hyperprolactinaemia. Norquetiapine (N-desalkyl quetiapine) is an active metabolite of quetiapine. It has a high affinity for the norepinephrine transporter and a partial agonist activity at the serotonin 5-HT1A receptor (Goldstein 2007). This profile might make quetiapine differ from other antipsychotics.

Why it is important to do this review

Since 1988, a newer generation of antipsychotic drugs has become available. These ‘atypical’ antipsychotics are defined as antipsychotic drugs with low propensity to induce extrapyramidal side effects (Kerwin 1994). Although atypical antipsychotics may cause less extrapyramidal adverse effects and movement disorders than typical antipsychotics, many of them are more likely to cause metabolic adverse effects. Moreover, the role of atypical antipsychotics in the treatment of schizophrenia is still under debate. Previous systematic reviews reported that there was no clear evidence that atypical antipsychotics were more effective than typical antipsychotics (Geddes 2000; Leucht 2008). Results from the two independent randomised controlled trials contradict previous trials comparing typical with atypical antipsychotics and the findings do question whether there is a meaningful difference between the old and new generation of drugs (Jones 2006; vs PERPHEN - L'rman 2005).

This systematic review aims to assess the evidence of efficacy and safety of quetiapine in comparison with typical antipsychotic drugs in the treatment of schizophrenia.

Objectives

To review the effects of quetiapine in comparison with typical antipsychotics in the treatment of schizophrenia and schizophrenia-like psychosis.

Methods

Criteria for considering studies for this review

Types of studies

We considered all relevant randomised controlled trials. We excluded quasi-randomised studies, such as those allocating by using alternate days of the week. Where trials were described as 'double-blind', but it was implied that the study was randomised and where the demographic details of each group's participants were similar, we included these trials and a sensitivity analysis was undertaken to the presence or absence of these data.

Randomised cross-over studies were eligible but only data up to the point of first cross-over because of the instability of the problem behaviours and the likely carry-over effects of all treatments (Elbourne 2002).

Types of participants

Participants included people with schizophrenia or other types of schizophrenia-like psychosis (for example, schizophreniform and schizoaffective disorders), irrespective of the diagnostic criteria used, age, ethnicity and sex. There is no clear evidence that the schizophrenia-like psychoses are caused by fundamentally different disease processes or require different treatment approaches (Carpenter 1994). Where a study described the participant group as suffering from 'serious mental illnesses' and did not give a particular diagnostic grouping, we included these trials. The exception to this rule was when the majority of those randomised clearly did not have a functional non-affective psychotic illness.

Types of interventions

1. Quetiapine: any oral form of application, any dose.

2. Typical antipsychotic drugs, that is, any other antipsychotics excluding amisulpride, sulpiride, zotepine, olanzapine, risperidone, sertindole, aripiprazole, ziprasidone and clozapine, at any dose.

Types of outcome measures

We grouped outcomes into the short term (up to 12 weeks), medium term (13-26 weeks) and long term (over 26 weeks).

Primary outcomes
Global state

No clinically important response as defined by the individual studies - for example, global impression less than much improved or less than 50% reduction on a rating scale.

Secondary outcomes
1. Leaving the studies early

Any reason, adverse events, inefficacy of treatment.

2. Relapse

Relapse (as defined by the individual studies).

3. Mental state (with particular reference to the positive and negative symptoms of schizophrenia)

3.1 No clinically important change in general mental state score
3.2 Average endpoint general mental state score
3.3 Average change in general mental state score
3.4 No clinically important change in specific symptoms (positive symptoms of schizophrenia, negative symptoms of schizophrenia)
3.5 Average endpoint specific symptom score
3.6 Average change in specific symptom score

4. General functioning

4.1 No clinically important change in general functioning
4.2 Average endpoint general functioning score
4.3 Average change in general functioning score

5. Quality of life/satisfaction with treatment

5.1 No clinically important change in general quality of life
5.2 Average endpoint general quality of life score
5.3 Average change in general quality of life score

6. Cognitive functioning

6.1 No clinically important change in overall cognitive functioning
6.2 Average endpoint of overall cognitive functioning score
6.3 Average change of overall cognitive functioning score

7. Service use

7.1 Number of participants hospitalised

8. Adverse effects

8.1 Number of participants with at least one adverse effect
8.2 Clinically important specific adverse effects (cardiac effects, death, movement disorders, prolactin increase and associated effects, sedation, seizures, weight gain, effects on white blood cell count)
8.3 Average endpoint in specific adverse effects
8.4 Average change in specific adverse effects

Search methods for identification of studies

No language restriction was applied within the limitations of the search tools.

Electronic searches

The Cochrane Schizophrenia Group Trials Register (November 2008 and March 2010) was searched with the following search strategy:

[(*quetiapine* or *seroquel* or *ICI-204636* or (*ICI* and *204636*) or *ICI204636* in title, abstract or index terms of REFERENCE) or (*quetiapine* in interventions of STUDY)]

The Cochrane Schizophrenia Group's Trials Register is based on regular searches of BIOSIS Inside; CENTRAL; CINAHL; EMBASE; MEDLINE and PsycINFO; the handsearching of relevant journals and conference proceedings, and searches of several key grey literature sources. A full description is given in the group's module.

Searching other resources

1. Reference searching

The reference lists of all retrieved articles, previous reviews and major text books of schizophrenia were examined for additional trials.  

2. Personal contact

We identified the authors of significant papers from authorship of trials and review articles found in the search. We contacted them, as well as other experts in the field, and asked for their knowledge of other studies, published or unpublished, relevant to the review.

3. Drug companies

We contacted the pharmaceutical company that manufactures quetiapine (Astra Zeneca) and requested relevant published and unpublished data. They provided a link to the studies that were carried out by the company including published and unpublished data. However, the data from the company proved to be the same as the data from the Cochrane Schizophrenia Group.

Data collection and analysis

Selection of studies

Review authors SJS and MS independently inspected citations identified from the search. We identified potentially relevant reports and ordered full papers for reassessment. Retrieved articles were assessed independently by STS and NM for inclusion according to the previously defined inclusion criteria. Any disagreement was resolved by consensus discussions with SJS. If it was impossible to resolve disagreements, we added these studies to those awaiting assessment and we contacted the authors of the papers for clarification. If there had been non-concurrence in trial selection, we would have reported this.

Data extraction and management

1. Extraction

SJS and BM independently extracted data from included studies. Any disagreement was discussed with MS, decisions documented and, if necessary, we contacted authors of studies for clarification. Due to language barrier, the data from Chinese studies were extracted by JX only.

2. Management

We extracted data onto standard, simple forms.

3. Scale-derived data

We included continuous data from rating scales only if: (a) the psychometric properties of the measuring instrument had been described in a peer-reviewed journal (Marshall 2000); (b) the measuring instrument was not written or modified by one of the trialists; (c) the measuring instrument was either (i) a self-report or (ii) completed by an independent rater or relative (not the therapist).

4. 'Summary of findings' table

We used the GRADE approach to interpret findings (Schünemann 2008) and used GRADE profiler (GRADE Profiler) to import data from RevMan 5.1 (Review Manager) to create a 'Summary of findings' table. This table provides outcome-specific information concerning the overall quality of evidence from each included study in the comparison, the magnitude of effect of the interventions examined, and the sum of available data on all outcomes we rated as important to patient-care and decision-making. We selected the following main outcomes for inclusion in the 'Summary of findings' table.

  • Global state

  • Leaving the study early

  • Mental state: positive, negative symptoms

  • Cognitive function

  • Adverse effects: extrapyramidal effects, prolactin level

Assessment of risk of bias in included studies

SJS and MS worked independently to assess risk of bias by using criteria described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) to assess trial quality. This set of criteria is based on evidence of associations between overestimate of effect and high risk of bias of the article such as sequence generation, allocation concealment, blinding, incomplete outcome data and selective reporting.

If the raters disagreed, the final rating was made by consensus discussion with BM. Where inadequate details of randomisation and other characteristics of trials were provided, we contacted the authors of the studies in order to obtain further information. Non-concurrence in quality assessment was reported, but if disputes arose as to which category a trial was to be allocated, again, resolution was made by discussion.

The level of risk of bias is noted in both the text of the review and in the Summary of findings for the main comparison.

Measures of treatment effect

1. Dichotomous data

Where possible, we made efforts to convert outcome measures to dichotomous data. This can be done by identifying cut-off points on rating scales and dividing participants accordingly into 'clinically improved' or 'not clinically improved'. It is generally assumed that if there had been a 50% reduction in a scale-derived score such as the Brief Psychiatric Rating Scale (BPRS, Overall 1962) or the Positive and Negative Syndrome Scale (PANSS, Kay 1986), this could be considered as a clinically significant response (Leucht 2005a; Leucht 2005b). If data based on these thresholds were not available, we used the primary cut-off presented by the original authors.

We calculated the risk ratio (RR) and its 95% confidence interval (CI) based on the random-effects model, as this takes into account any differences between studies, even if there is no statistically significant heterogeneity. It has been shown that RR is more intuitive (Boissel 1999) than odds ratios and that odds ratios tend to be interpreted as RR by clinicians (Deeks 2000). This misinterpretation then leads to an overestimate of the impression of the effect. When the overall results were significant, we calculated the number needed to treat to provide benefit (NNTB) and the number needed to treat to induce harm (NNTH) as the inverse of the risk difference.

2. Continuous data
2.1 Summary statistic

For continuous outcomes we estimated a mean difference (MD) between groups. Mean differences were based on the random-effects model as this takes into account any differences between studies even if there is no statistically significant heterogeneity. We did not calculate standardised mean differences (SMD) measures.

2.2 Endpoint versus change data

Since there is no principal statistical reason why endpoint and change data should measure different effects (Higgins 2011), we used scale endpoint data, which is easier to interpret from a clinical point of view. If endpoint data were not available, we used changed data.

2.3 Skewed data

Continuous data on clinical and social outcomes are often not normally distributed. To avoid the pitfall of applying parametric tests to non-parametric data, we applied the following standards to all data before inclusion:

(a) standard deviations and means are reported in the paper or obtainable from the authors;
(b) when a scale starts from the finite number zero, the standard deviation, when multiplied by two, is less than the mean (as otherwise the mean is unlikely to be an appropriate measure of the centre of the distribution, (Altman 1996);
(c) if a scale starts from a positive value (such as PANSS which can have values from 30 to 210), the calculation described above was modified to take the scale starting point into account. In these cases skew is present if 2 SD > (S-S min), where S is the mean score and S min is the minimum score. Endpoint scores on scales often have a finite start and end point and these rules can be applied. When continuous data are presented on a scale that includes a possibility of negative values (such as change data), it is difficult to tell whether data are skewed or not.

We planned to enter skewed data from studies of less than 200 participants in 'other data tables' rather than into analyses. However, we found that most of the data from included studies were skewed (e.g. the data from all the studies investigated PANSS positive symptoms) and excluding all studies on the basis of estimates of the normal distribution would lead to selection bias. We therefore included all studies in the primary analysis and excluded the skewed data in the sensitivity analysis (see Differences between protocol and review).

2.4 Data synthesis

When standard errors instead of standard deviations were presented, the former were converted to standard deviations. If standard deviations were not reported and could not be calculated from available data, authors were asked to supply the data. In the absence of data from authors, the mean standard deviations from other studies were used.

2.5 Multiple doses

When a study investigated a number of fixed doses of quetiapine, we used scores from the highest dose group.

Unit of analysis issues

1. Cluster trials

Studies increasingly employ 'cluster randomisation' (such as randomisation by clinician or practice) but analysis and pooling of clustered data poses problems. Firstly, authors often fail to account for intraclass correlation in clustered studies, leading to a 'unit of analysis' error (Divine 1992) whereby P values are spuriously low, confidence intervals unduly narrow and statistical significance overestimated. This can lead to type I errors or a false positive (Bland 1997; Gulliford 1999).

Where clustering was not accounted for in primary studies, we presented the data in a table, with a (*) symbol to indicate the presence of a probable unit of analysis error. In subsequent versions of this review, we will seek to contact first authors of studies to obtain intraclass correlation coefficients (ICCs) of their clustered data and to adjust for this using accepted method (Gulliford 1999). If we had found that clustering was incorporated into the analysis of primary studies, we would have presented these data as if from a non-cluster randomised study, but adjusted for the clustering effect.

Binary data as presented in a report were divided by a 'design effect'. This was calculated using the mean number of participants per cluster (m) and the ICC [Design effect = 1+(m-1)*ICC] (Donner 2002). If the ICC was not reported, it was assumed to be 0.1 (Ukoumunne 1999). This assumption may be too high and, had this instance occured, we had planned to see if taking an ICC of 0.01 would make any substantive difference for the primary outcome. If it had not, we would have used 0.01 in preference across outcomes.

If we had found that cluster studies were appropriately analysed taking into account ICCs and relevant data documented in the report, we would have synthesised these with other studies using the generic inverse variance technique.

2. Cross-over trials

A major concern of cross-over trials is the carry-over effect. It occurs if an effect (e.g. pharmacological, physiological or psychological) of the treatment in the first phase is carried over to the second phase. As a consequence on entry to the second phase the participants can differ systematically from their initial state despite a wash-out phase. For the same reason cross-over trials are not appropriate if the condition of interest is unstable (Elbourne 2002). As both effects are very likely in schizophrenia, we would only have used data of the first phase of cross-over studies.

3. Studies with multiple treatment groups

If we had found any study with more than two treatment groups, we would have presented the additional treatment groups in additional relevant comparisons. Data were not double counted. Where the additional treatment groups were not relevant, these data were not reproduced.

Dealing with missing data

At some degree of loss of follow-up, data must lose credibility (Xia 2009), therefore, we planned to exclude studies with an attrition rate over 40%. However, many of the main studies including vs HLP - Arvanitis 1997, vs FLUPHEN - Conley 2005, vs HLP - Fl'hacker 2005, vs PERPHEN - L'rman 2005, vs HLP - Purdon 2001 and vs HLP - Velligan 2002) reported high attrition rates over 40%. Moreover, it is still unclear what degree of attrition leads to a high degree of bias (Komossa 2010). We, therefore, did not exclude these studies on the basis of the attrition rate, but we carried out sensitivity analyses of the main mental state outcomes excluding the studies with high attrition rates. We also have addressed the attrition problems as well as the use of intention-to-treat (ITT) in the 'Risk of bias' table, the results and discussion sections.

Intention-to-treat was used when available. When these data were not clearly described, data were presented on a 'once-randomised-always-analyse' basis, assuming an ITT analysis. We anticipated that in some studies, in order to do an ITT analysis, the method of last-observation-carried-forward (LOCF) would be employed within the study report. As with all methods of imputation to deal with missing data, LOCF introduces uncertainty about the reliability of the results. Therefore, where LOCF data have been used in the analysis, it was indicated in the review. Sensitivity analyses excluding studies with the use of ITT were undertaken to test how prone the primary outcomes were to change when 'completer' data only were included in the analyses.

Assessment of heterogeneity

1. Clinical heterogeneity

We considered all included studies hoping to use all studies together. Where clear unforeseen issues were apparent that might have added obvious clinical heterogeneity, we noted these issues, considered them in analyses and undertook sensitivity analyses for the primary outcome.

2. Statistical
2.1 Visual inspection

We visually inspected graphs to investigate the possibility of statistical heterogeneity.

2.2 Employing the I2 statistic

Heterogeneity between studies was investigated by considering the I2 method alongside the Chi2 'P' value. The I2 provides an estimate of the percentage of inconsistency thought to be due to chance (Higgins 2003). The importance of the observed value of I2 depends on i. magnitude and direction of effects and ii. strength of evidence for heterogeneity (e.g. 'P' value from Chi2 test, or a confidence interval for I2).

An I2 estimate greater than or equal to 50% accompanied by a statistically significant Chi2 statistic, was interpreted as evidence of substantial levels of heterogeneity (Section 9.5.2 - Higgins 2011) and reasons for heterogeneity were explored. If the inconsistency was high and the clear reasons were found, data were presented separately.

Assessment of reporting biases

Reporting biases arise when the dissemination of research findings is influenced by the nature and direction of results (Egger 1997). These are described in section 10.1 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We are aware that funnel plots may be useful in investigating small-study effects but are of limited power to detect such effects when there are few studies. We did not use funnel plots for outcomes where there were 10 or fewer studies, or where all studies were of similar sizes. In other cases, where funnel plots were possible, we sought statistical advice in their interpretation.

Data synthesis

Where possible, we employed a random-effects model for analyses. We understand that there is no closed argument for preference for use of fixed-effect or random-effects models. The random-effects method incorporates an assumption that the different studies are estimating different, yet related, intervention effects. This does seem true to us, however, random-effects does put added weight onto the smaller of the studies - those trials that are most vulnerable to bias.

Subgroup analysis and investigation of heterogeneity

If data were clearly heterogeneous, we checked that data were correctly extracted and entered and that we had made no unit-of-analysis errors. If high levels of heterogeneity remained, we did not undertake a meta-analysis at this point as if there is considerable variation in results, and particularly if there is inconsistency in the direction of effect, it may be misleading to quote an average value for the intervention effect. We planned to explore heterogeneity. We pre-specified no characteristics of studies that may be associated with heterogeneity except for methodological quality. If no clear association could be shown by sorting studies by quality of methods, we continued to investigate for other reasons for the heterogeneity. Had we identified another characteristic of the studies by the investigation of heterogeneity, perhaps some clinical heterogeneity not hitherto predicted - but plausible causes of heterogeneity - we planned to discuss these post-hoc reasons, investigate the sensitivity of the estimate of the effect size for the primary outcome to inclusion and exclusion of these causes, and analyse and present the data. However, if the heterogeneity was substantially unaffected by any investigation and no reasons for the heterogeneity were apparent, we planned to present the final data without a meta-analysis. Subgroup analysis was not carried out in this review.

Sensitivity analysis

We planned sensitivity analyses a priori for examining the change in the robustness of the sensitivity to including studies with implied randomisation (see Criteria for considering studies for this review: Types of studies), skewed and non-skewed data, inappropriate comparator doses of drug and different clinical groups. However, we did not find any difference in clinical groups, therefore, the sensitivity analyses on this matter was not done. We also added the sensitivity analyses of the main mental state outcomes excluding the studies with high attrition rate.

If inclusion of studies with implied randomisation made no substantive difference to the primary outcome they were left in the final analyses. For outcomes with both skewed data and non-skewed data, we investigated the effect of combining all data together and if no substantive difference was noted then the potentially skewed data were left in the analyses. A recent review showed that some of the comparisons of antipsychotics may have been biased by using inappropriate comparator dose ranges (Heres 2006). The inappropriate dose ranges were defined as the ranges not within the range recommended in the American Psychiatric Association Practice Guideline for the treatment of patients with Schizophrenia, second edition (APA 2004).

Drugmg/day
Quetiapine300-800
Typical
Chlorpromazine300-1000
Fluphenazine5-20
Haloperidol5-20
Loxapine30-100
Mesoridazine150-400
Molindone30-100
Perphenazine16-64
Thioridazine300-800
Thiothixene15-50
Trifluoperazine15-50

If we had found the studies with implied randomisation, skewed and non-skewed data, inappropriate comparator doses of drug and different clinical groups, we would have analysed whether the exclusion of these studies changed the results of the primary outcome and the general mental state.

Results

Description of studies

See also: Characteristics of included studies; Characteristics of excluded studies. In this review we use the style of including the comparison compound as part of the study tag. For example the study 'vs CPZ - Ai 2007' involves quetiapine compared with chlorpromazine. Other abbreviations used are HLP - haloperidol, FLUPHEN - fluphenazine and PERPHEN - perphenazine. In our opinion, this style conveys more information in each graph as simply using the name of the first author leaves unclear which comparison is being used.

Results of the search

The overall search strategy yielded 830 reports (666 in 2008 and 164 in 2010) of which 65 were closely inspected (Figure 1).

Figure 1.

PRISMA flow diagram.

Included studies

Forty-three studies with 7217 participants met the inclusion criteria. Twelve studies were sponsored by the pharmaceutical companies developing quetiapine.

1. Length of trials

Thirty-five studies were short term with a duration range of four to 12 weeks. Five studies were medium term and three studies were long term.

2. Settings

Twenty-eight studies were conducted in China. Of the 43 included studies, 11 studies were conducted in an inpatient or outpatient setting, 20 studies were conducted in inpatient settings and three studies were conducted in outpatient settings. Two studies were conducted in communities and hospitals. Seven studies did not report the setting.

3. Participants

Twenty Chinese studies included participants diagnosed by using the Chinese Classification of Mental Disorders, third version (CCMD-3) and five Chinese studies used CCMD-2. Eleven studies included participants diagnosed by using the Diagnostic and Statistical Manual Fourth revision (DSM-IV) and four studies used the third edition, revised (DSM-III-R). Two studies diagnosed participants by using the International Classification of Diseases Version 10 (ICD-10) and vs CPZ - Zhou 2004 used both ICD-10 and CCMD-3. Three studies included only acutely ill participants (vs HLP - Copolov 2000; vs CPZ - Guo 2003; vs HLP - Taneli 2003). Two studies included only participants with a first episode schizophrenia (vs HLP - Fl'hacker 2005; vs CPZ - Hu 2003). One study focused on participants with full or partial remission (vs HLP - Velligan 2002) and one study focused on treatment-resistant participants (vs HLP - Emsley 1999).

4. Study size

vs PERPHEN - L'rman 2005 was the largest study with 1493 participants and vs HLP - Purdon 2001 was the smallest with only 25. Four studies had less than 50 participants but three studies randomised more than 400 people.

5. Interventions
5.1 Quetiapine

Four studies used fixed dosing (vs HLP - Arvanitis 1997; vs HLP - Atmaca 2002; vs HLP - Emsley 1999; vs HLP - Velligan 2002) while others used flexible regimens. Overall, quetiapine was given in a dose range of 50-800 mg. However, vs CPZ - Zhong 2005 and vs HLP - McCue 2006 limited the upper dose range at 900 mg/day and 1200 mg/day respectively.

5.2 Typical antipsychotic drugs

The comparator typical antipsychotics were chlorpromazine, haloperidol and perphenazine. Again, four studies used fixed dosing (vs HLP - Arvanitis 1997; vs HLP - Atmaca 2002; vs HLP - Emsley 1999; vs HLP - Velligan 2002) while others used a flexible regimen. Some studies included treatment arms of participants given other atypical antipsychotics but these results were not reported in this review.

6. Outcomes
6.1 Leaving the study early

The number of participants leaving the studies early was reported for the categories 'any reason', 'adverse events' and 'lack of efficacy'.

6.2 Outcome scales

Details of scales that provided usable data are shown below.

6.2.1 Global state scales

6.2.1.1 Clinical Global Impression Scale - CGI (Guy 1976)
This is used to assess both severity of illness and clinical improvement, by comparing the conditions of the person standardised against other people with the same diagnosis. A seven-point scoring system is usually used with low scores showing decreased severity and/or overall improvement.

6.2.2 Mental state scales

6.2.2.1 Positive and Negative Syndrome Scale - PANSS (Kay 1986)
This schizophrenia scale has 30 items, each of which can be defined on a seven-point scoring system varying from one (absent) to seven (extreme). It can be divided into three subscales for measuring the severity of general psychopathology, positive symptoms (PANSS-P) and negative symptoms (PANSS-N). A low score indicates lesser severity.

6.2.2.2 Brief Psychiatric Rating Scale - BPRS (Overall 1962)
This is used to assess the severity of abnormal mental state. The original scale has 16 items, but a revised 18-item scale is commonly used. Each item is defined on a seven-point scale varying from 'not present' to 'extremely severe', scoring from zero to six or one to seven. Scores can range from zero to 126 with high scores indicating more severe symptoms.

6.2.2.3 Scale for the Assessment of Negative Symptoms - SANS (Andreasen 1989)
This six-point scale gives a global rating of the following negative symptoms: alogia, affective blunting, avolition-apathy, anhedonia-asociality and attention impairment. Assessments are made on a six-point scale from zero (not at all) to five (severe). Higher scores indicate more severe symptoms.

6.2.3 General functioning

6.2.3.1 Global Assessment of Functioning - GAF (APA 1994)
A rating scale for a patients´ overall capacity of psychosocial functioning, scoring from one to 100. Higher scores indicate a higher level of functioning.

6.2.4 Quality of life/satisfaction with treatment

6.2.4.1 Manchester Short Assessment of Quality of Life - MANSA (Priebe 1999)
A rating scale to assess quality of life focusing on satisfaction with life as a whole and with life domains. Higher scores indicate less impairment.

6.2.4.2 Quality of Life Scale - QLS (Carpenter 1984)
This semi-structured interview is administered and rated by trained clinicians. It contains 21 items rated on a seven point scale based on the interviewers´ judgment of patient functioning. A total QLS and four subscale scores are calculated, with higher scores indicate less impairment.

6.2.4.3 The Short Form (36) Health Survey - SF-36 (Ware 1992)
This self-reporting measure consists of eight scaled scores. Each scale is directly transformed into a zero to 100 scale on the assumption that each question carries equal weight.

6.2.5 Adverse effects scales

6.2.5.1 Abnormal Involuntary Movement Scale - AIMS (Guy 1976)
This has been used to assess tardive dyskinesia, a long-term, drug-induced movement disorder and short-term movement disorders such as tremor.

6.2.5.2 Extrapyramidal Symptom Rating Scale - ESRS (Chouinard 1980)
This is a questionnaire relating to parkinsonian symptoms (nine items), a physician’s examination for parkinsonism and dyskinetic movements (eight items), and a clinical global impression of tardive dyskinesia. High scores indicate severe levels of movement
disorder.

6.2.5.3 Simpson Angus Scale - SAS (Simpson 1970)
This is a 10-item scale, with a scoring system of zero to four for each item, measures drug-induced parkinsonism, a short-term drug-induced movement disorder. A low score indicates low levels of Parkinsonism.

6.2.5.4 Treatment Emergent Symptom Scale - TESS (Guy 1976)
The scale measures adverse events. A low score indicates low levels of adverse events.

6.3 Other adverse effects

Many adverse effects were reported as continuous variables for QTc prolongation (ms), cholesterol level (mg/dL), glucose levels (mg/dL), prolactin level (ng/mL) and weight (kg). Other adverse events were reported in a dichotomous manner in terms of the number of people with a given effect.

6.4 Service use

Service use was described as the number of patients re-hospitalised during the trial.

Excluded studies

Twenty-two studies had to be excluded for the following reasons:

Four were not randomised, 15 reported no usable data, one had high risk of bias from being an open-label study, with incomplete outcome data, one used combined antipsychotics and one was a pooled analysis rather than a trial.

Risk of bias in included studies

For details please refer to 'Risk of bias' tables for each study and Figure 2 and Figure 3.

Figure 2.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Figure 3.

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.

Allocation

All of the included studies were described as randomised. Only nine studies gave further information on the randomisation, three used random number tables (vs CPZ - Chen 2001; vs CPZ - Peng 2006; vs CPZ - Zhou 2003), three computer-generated randomisation (vs HLP - Fl'hacker 2005; vs HLP - McCue 2006; vs CPZ - Zhang 2003) and two using coin tossing (vs CPZ - Guo 2003;vs CPZ - Zhang 2002). For all other studies, it was unclear whether the allocation strategies were appropriate.

Blinding

Four of the included studies were 'single-blind', 12 studies were 'double-blind'. One study used identical capsules for blinding (vs PERPHEN - L'rman 2005). The other trials did not provide any information on the blinding procedure. No study examined the effectiveness of blinding. Seven studies were open-label trials. For the other studies, blinding issues were not reported.

Incomplete outcome data

Twenty-four studies reported the number of participants leaving the studies early for any reason. A major problem was the high attrition in which six studies had more than 40% (vs HLP - Arvanitis 1997; vs FLUPHEN - Conley 2005; vs HLP - Fl'hacker 2005; vs PERPHEN - L'rman 2005; vs HLP - Purdon 2001; vs HLP - Velligan 2002). In most studies the last-observation-carried-forward (LOCF) method was used to compensate for attrition. The sensitivity analyses excluding studies with high attrition rates were added.

Selective reporting

Selective reporting was judged to be present in 16of the 43 included studies (37%). The main reason was the incomplete reporting of predefined outcomes.

Other potential sources of bias

Twelve studies were sponsored by pharmaceutical companies (vs HLP - Arvanitis 1997; vs CPZ - AstraZeneca 2005; vs FLUPHEN - Conley 2005; vs HLP - Copolov 2000; vs HLP - Emsley 1999; vs HLP - Emsley 2004; vs HLP - Fl'hacker 2005; vs CPZ - Link 1997; vs HLP - Murasaki 1999; vs HLP - Purdon 2001; vs HLP - Taneli 2003; vs HLP - Velligan 2002). Pharmaceutical companies sometimes highlight the benefits of the agent and mentioned less on its disadvantages (Heres 2006). One study included only female participants; therefore, gender bias could not be excluded (vs HLP - Atmaca 2002).

Effects of interventions

See: Summary of findings for the main comparison Quetiapine compared to Typical antipsychotics for schizophrenia

1. Comparison 1. Quetiapine versus typical antipsychotics

Forty-three studies met the inclusion criteria for this comparison.

1.1 Global state: 2. No clinical improvement

There was no significant difference in short-term data (15 RCTs, n = 1479, risk ratio (RR) 1.05 confidence interval (CI) 0.81 to 1.35) but one medium-term study found a significant difference, favouring the treatment group (1 RCT, n = 128, RR 0.16 CI 0.05 to 0.51, Analysis 1.1).

1.2 Global state: 1a. CGI: Average CGI-S (high = poor)

There was no significant difference in short-term endpoint data (5 RCTs, n = 347, mean difference (MD) 0.22 CI -0.04 to 0.49) and long-term endpoint data (1 RCT, n = 207, MD -0.10 CI -0.93 to 0.73). However, there was a significant difference, favouring the control group in short-term change data (2 RCTs, n = 699, MD 0.20 CI 0.04 to 0.36). The data were heterogenous (I2 = 64%). The heterogeneity was more likely due to differences in degree of change in CGI-S than directions of effect (Analysis 1.2).

1.3 Global state: 1b. CGI: Average endpoint CGI-I (high = poor)

There was a significant difference, favouring the control group in short-term (4 RCTs, n = 496, MD 0.39 CI 0.30 to 0.47) but not in a medium-term studies (1 RCT, n = 11, MD -0.20 CI -1.08 to 0.68, Analysis 1.3).

1.4 Leaving the study early: Any reason

The number of participants who left the studies early due to any reason were similar (36.5% in the treatment group and 36.9% in the control group). There was a significant difference, favouring the treatment group in medium-term (3 RCTs, n = 203, RR 0.72 CI 0.52 to 0.99) but not in short-term (17 RCTs, n = 2535, RR 0.88 CI 0.78 to 1.00) and long-term studies (3 RCTs, n = 838, RR 0.99 CI 0.77 to 1.28, Analysis 1.4).

1.5 Leaving the study early: Adverse events

Fewer participants in the treatment group (6.6%) compared with the control group (11.6%) left the studies early due to adverse events. There was a significant difference, favouring the treatment group in short-term studies (11 RCTs, n = 2044, RR 0.51 CI 0.31 to 0.84) and one medium-term study (1 RCT, n = 128, RR 0.03 CI 0.00 to 0.53). However, there was no significant difference in long-term study (3 RCTs, n = 838, RR 0.46 CI 0.12 to 1.78) and the long-term data were heterogenous (Analysis 1.5).

1.6 Leaving the study early: Inefficacy

There was no significant difference in both short-term (3 RCTs, n = 842, RR 1.20 CI 0.87 to 1.66) and long-term studies (6 RCTs, n = 1221, RR 1.41 CI 0.94 to 2.11, Analysis 1.6).

1.7 Mental state: 1a. General - average score (PANSS total, high = poor)

There was no significant difference of endpoint score in short-term (21 RCTs, n = 2055, MD 0.84 CI -1.46 to 3.15), medium-term (3 RCTs, n = 180, MD -7.77 CI -15.69 to 0.14) and long-term studies (2 RCTs, n = 252, MD -0.90 CI -4.94 to 3.15). However, there was a significant difference, favouring the control group, in short-term change score (2 RCTs, n = 565, MD 3.59 CI 0.09 to 7.10). Heterogeneity might be due to directions of effect and one outlier (vs CPZ - Tian 2006). However, excluding this study still showed no significant difference in short-term data (21 RCTs, n = 2052, MD -0.44 CI -0.44 to 0.68). Overall, there was no significant difference between groups on PANSS total score (27 RCTs, n = 3052, MD 0.09 CI -2.14 to 2.31, Analysis 1.7).

1.8 Mental state: 1b General - average score - short term (BPRS total, high = poor)

There was no significant difference in endpoint score (6 RCTs, n = 666, MD -0.24 CI -1.66 to 1.17) but there was a significant difference, favouring the control group, in change score (2 RCTs, n = 359, MD 3.49 CI 0.90 to 6.08). Overall, there was no significant difference between groups on BPRS total score (8 RCTs, n = 1025, MD 0.71 CI -0.77 to 2.20, Analysis 1.8).

1.9 Mental state: 2a. Positive symptoms - average endpoint score (PANSS positive subscore, high = poor)

There was no significant difference in short-term (19 RCTs, n = 1801, MD 0.19 CI -0.25 to 0.62), medium-term (2 RCTs, n = 88, MD -1.57 CI -3.41 to 0.26) and long-term studies (1 RCT, n = 45, MD -1.30 CI -3.12 to 0.52, Analysis 1.9).

1.10 Mental state: 2b. Positive symptoms - average score - short term (BPRS positive subscore, high = poor)

There was no significant difference in endpoint score (2 RCTs, n = 128, MD 0.12 CI -0.42 to 0.65) but there was a significant difference, favouring the control group (1 RCT, n = 253, MD 1.34 CI 0.27 to 2.41). Overall, there was no significant difference between groups on BPRS positive subscore (3 RCTs, n = 381, MD 0.63 CI -0.37 to 1.62, Analysis 1.10).

1.11 Mental state: 3a. Negative symptoms - average endpoint score (PANSS negative subscore, high = poor)

There was no significant difference in short-term (19 RCTs, n = 1801, MD -0.81. CI -1.63 to 0.01) and long-term studies (1 RCT, n = 45, MD 1.20 CI -2.09 to 4.49) but there was a significant difference, favouring the treatment group, in medium-term studies (2 RCTs, n=88, MD -2.45 CI -4.64 to -0.27). Overall, there was a significant difference, favouring the treatment group, on PANSS negative subscore (22 RCTs, n = 1934, MD -0.82 CI -1.59 to -0.04, Analysis 1.11), but this result was highly heterogeneous and driven by two small outlier studies with high effect sizes. Without these two studies, there was no heterogeneity and no statistically significant difference between the treatment and the control group.

1.12 Mental state: 3b. Negative symptoms - average score - short term (BPRS negative subscore, high = poor)

There was no significant difference in endpoint score (1 RCT, n = 25, MD 0.11 CI -2.01 to 2.23) but there was a significant difference, favouring the treatment group in change score (1 RCT, n = 253, MD -0.75 CI -1.42 to -0.08). Overall, there was a significant difference, favouring the treatment group, in BPRS negative subscore (2 RCTs, n = 278, MD -0.67 CI -1.31 to -0.04, Analysis 1.12).

1.13 Mental state: 3c. Negative symptoms - average endpoint score - short term (SANS total, high = poor)

There was a significant difference, favouring the control group, in endpoint score (1 RCT, n = 103, MD 1.80 CI 0.24 to 3.36) but no significant difference in change score (1 RCT, n = 228, MD -0.26 CI -1.08 to 0.56). Overall, there was no significant difference between the groups on SANS total score (2 RCTs, n = 331, MD 0.66 CI -1.35 to 2.66, Analysis 1.13).

1.14 Mental state: 4. General psychopathology - average endpoint score (PANSS general psychopathology subscore, high = poor)

There was no significant difference in short-term (15 RCTs, n = 1472, MD -0.17 CI -0.81 to 0.47), medium-term (2 RCTs, n = 52, MD 0.47 CI -3.30 to 4.24) and long-term studies (1 RCT, n = 45, MD -2.20 CI -6.02 to 1.62, Analysis 1.14).

1.15 General functioning: General - average endpoint score - long term (GAF total score, low = poor)

There was no significant difference (1 RCT, n = 207, MD -0.10 CI -9.80 to 9.60, Analysis 1.15).

1.16 Quality of life: 1a. General - average endpoint score - long term (SF-36, low = poor)

There was a significant difference, favouring the treatment group (1 RCT, n = 84, MD 13.76 CI 5.95 to 21.57, Analysis 1.16).

1.17 Quality of life: 1b. General - average endpoint score - long term (MANSA total score, low = poor)

There was no significant difference (1 RCT, n = 207, MD 0.00 CI -1.38 to 1.38, Analysis 1.17).

1.18 Quality of life: 1c. General - average change score - long term (QLS, high = poor)

There was no significant difference (1 RCT, n = 156, MD 0.10 CI -0.23 to 0.43, Analysis 1.18).

1.19 Cognitive function: 1a General - average change score - long term (Composite score, high = poor)

There was no significant difference (2 RCTs, n = 407, MD 0.00 CI -0.19 to 0.20, Analysis 1.19).

1.20 Cognitive function: 1b. General - average endpoint scores as defined by the original studies (low = poor)

There was a significant difference, favouring the treatment group in short-term (WAIS) (1 RCT, n = 91, MD 8.30 CI 1.64 to 14.96) but not in long-term studies (2 RCTs, n = 51, MD 0.82 CI -0.77 to 2.41, Analysis 1.20). The heterogeneity of long-term data might be due to the differences of cognitive scales used.

1.21 Service use: number of participants re-hospitalised - long term

There was no significant difference (2 RCTs, n = 722, RR 1.23 CI 0.90 to 1.68, Analysis 1.21).

1.22 Adverse effects: 1. General - at least one adverse effect

There was a significant difference, favouring the treatment group in short-term (7 RCTs, n = 1180, RR 0.75 CI 0.62 to 0.91, NNTH 9, CI 6 to 16) but not in long-term studies (2 RCTs, n = 805, RR 0.82 CI 0.52 to 1.29). The heterogeneity of short-term data might be due to an outlier (vs HLP - Emsley 1999), which showed a trend in favour of control. Excluding this study still showed a significant difference, favouring the treatment group, in short-term data (6 RCTs, n = 892, RR 0.73 CI 0.61 to 0.87). Overall, there was a significant difference, favouring the treatment group, in having at least one adverse effect (9 RCTs, n = 1985, RR 0.76 CI 0.64 to 0.90, Analysis 1.22).

1.23 Adverse effects: 2. Death

There was no significant difference of suicide attempt (1 RCT, n = 598, RR 0.77 CI 0.05 to 12.32) or suicide in one long-term study (1 RCT, n = 598, RR 0.52 CI 0.09 to 3.07) and death in one short-term study term (1 RCT, n = 260, RR 3.00 CI 0.12 to 72.97, Analysis 1.23).

1.24 Adverse effects: 3a. Cardiac effects - QTc prolongation - long term

There was no significant difference (1 RCT, n = 41, RR 1.73 CI 0.17 to 17.59, Analysis 1.24).

1.25 Adverse effects: 3b. Cardiac effects - abnormal ECG - short term

There was a significant difference, favouring the treatment group (2 RCTs, n = 165, RR 0.38 CI 0.16 to 0.92, NNTH 8, CI 4 to 55, Analysis 1.25).

1.26 Adverse effects: 3c. Cardiac effects - orthostatic hypotension

There was no significant difference in short-term studies (4 RCTs, n = 795, RR 0.45 CI 0.13 to 1.54) and one long-term study (1 RCT, n = 598, RR 1.01 CI 0.64 to 1.60). The heterogeneity of short-term data might be due to an outlier (vs HLP - Copolov 2000) which showed a trend in favour of control. Excluding this study revealed a significant difference in favour of quetiapine in short-term data (3 RCTs, n = 347, RR 0.25 CI 0.12 to 0.55, Analysis 1.26).

1.27 Adverse effects: 3d. Cardiac effects - low blood pressure - short term

There was no significant difference (6 RCTs, n = 572, RR 0.69 CI 0.41 to 1.18, Analysis 1.27).

1.28 Adverse effects: 3e. Cardiac effects - tachycardia - short term

There was no significant difference (8 RCTs, n = 814, RR 0.70 CI 0.36 to 1.34, Analysis 1.28).

1.29 Adverse effects: 4. Central nervous system - sedation

There was no significant difference in short-term studies (13 RCTs, n = 1700, RR 0.69 CI 0.37 to 1.30) and one long-term study (1 RCT, n = 598, RR 1.08 CI 0.84 to 1.39, Analysis 1.29). The heterogeneity of short term data might be due to various types of typical antipsychotics in control group.

1.30 Adverse effects: 5a. Extrapyramidal effects - overall - short term

There was a significant difference, favouring the treatment group (8 RCTs, n = 1095 RR 0.17 CI 0.09 to 0.32, NNTH 3, CI 3 to 3, Analysis 1.30).

1.31 Adverse effects: 5b. Extrapyramidal effects - akathisia

There were significant differences, favouring the treatment group, in both short-term (15 RCTs, n = 2059, RR 0.24 CI 0.17 to 0.35, NNTH 6, CI 5 to 8) and long-term studies (1 RCT, n = 158, RR 0.50 CI 0.25 to 0.98, NNTH 8, CI 4 to 153, Analysis 1.31).

1.32 Adverse effects: 5c. Extrapyramidal effects - parkinsonism

There were significant differences, favouring the treatment group, in both short-term (2 RCTs, n = 185, RR 0.13 CI 0.04 to 0.48, NNTH 6, CI 4 to 11) and long-term studies (1 RCT, n = 158, RR 0.31 CI 0.15 to 0.62, NNTH 5, CI 3 to 10, Analysis 1.32).

1.33 Adverse effects: 5d. Extrapyramidal effects - dystonia

There was a significant difference, favouring the treatment group, in short-term studies (4 RCTs, n = 834, RR 0.13 CI 0.04 to 0.51, NNTH 21, CI 15 to 38) but not in one long-term study (1 RCT, n = 158, RR 0.86 CI 0.05 to 13.49). Overall, there was a significant difference, favouring the treatment group (5 RCTs, n = 992, RR 0.19 CI 0.06 to 0.64, Analysis 1.33).

1.34 Adverse effects: 5e. Extrapyramidal effects - tremor - short term

There was a significant difference, favouring the treatment group (12 RCTs, n = 1641, RR 0.33 CI 0.23 to 0.47, NNTH 8, CI 7 to 12, Analysis 1.34).

1.35 Adverse effects: 5f. Extrapyramidal effects - scale measured (dichotomous data) - long term
1.35.1 Extrapyramidal symptoms: Simpson-Angus Scale (SAS >/=1)

There was no significant difference (1 RCT, n = 541, RR 0.65 CI 0.31 to 1.37).

1.35.2 Abnormal movement: Abnormal Involuntary Movement Scale (AIMS>/=2)

There was no significant difference (1 RCT, n = 473, RR 0.73 CI 0.48 to 1.14).

1.35.3 Akathisia: Barnes Akathisia Scale (BARS >/=3)

There was no significant difference (1 RCT, n = 546, RR 0.79 CI 0.40 to 1.55, Analysis 1.35).

1.36 Adverse effects: 5g. Extrapyramidal effects - scale measured (continuous data, high = poor)
1.36.1 Average endpoint score - Extrapyramidal symptoms (ESRS) - short term

There was a significant difference, favouring the treatment group (1 RCT, n = 35, MD -4.55 CI -6.58 to -2.52).

1.36.2 Average endpoint score - Abnormal Involuntary Movement Scale (AIMS) - medium term

There was no significant difference (1 RCT, n = 11, MD 0.30 CI -0.83 to 1.43).

1.36.3 Average endpoint score - Simpson-Angus Scale (SAS) - medium term

There was no significant difference (1 RCT, n = 11, MD 1.10 CI -1.94 to 4.14).

1.36.4 Average endpoint score - Treatment Emergent Symptom Scale (TESS) - short term

There was no significant difference (2 RCTs, n = 158, MD -1.09 CI -2.51 to 0.33).

1.36.5 Average change score - Treatment Emergent Symptom Scale (TESS) - short term

There was no significant difference (1 RCT, n = 117, MD -1.03 CI -2.49 to 0.43).

1.36.6 Average endpoint score - Treatment Emergent Symptom Scale (TESS) - medium term

There was no significant difference (1 RCT, n = 41, MD -2.10 CI -4.55 to 0.35, Analysis 1.36).

1.37 Adverse effects: 6a. Prolactin associated side effects
1.37.1 Gynaecomastia, galactorrhoea - long term

There was no significant difference (1 RCT, n = 598, RR 1.16 CI 0.33 to 4.07).

1.37.2 Menstrual irregularities - long term

There was no significant difference (1 RCT, n = 598, RR 0.55 CI 0.18 to 1.72, Analysis 1.37).

1.38 Adverse effects: 6b. Prolactin - Hyperprolactinemia

There was a significant difference, favouring the treatment group, in two short-term studies (2 RCTs, n = 165, RR 0.08 CI 0.01 to 0.60, NNTH 8, CI 5 to 17) but not in one long-term study (1 RCT, n = 64, RR 0.91 CI 0.51 to 1.62, Analysis 1.38).

1.39 Adverse effects: 6c. Prolactin level - average level in ng/mL

There was a significant difference, favouring the treatment group, in short-term endpoint level (1 RCT, n = 35, MD -15.67 CI -21.00 to -10.34), short-term change level (1 RCT, n = 356, MD -22.43 CI -23.20 to -21.66), long-term endpoint level (1 RCT, n = 45, MD -16.40 CI -23.83 to -8.97) and long-term change level (1 RCTs, n = 598, MD -9.70 CI -14.02 to -5.38). The heterogeneity was more likely to be due to differences in degree of prolactin decrease than directions of effect. Overall, there was a significant difference, favouring the treatment group, in prolactin level (4 RCTs, n = 1034, MD -16.20 CI -23.34 to -9.07, Analysis 1.39).

1.40 Adverse effects: 7a. Weight gain (as defined by the original studies)

There was a significant difference, favouring the treatment group in short-term (9 RCTs, n = 866, RR 0.52 CI 0.34 to 0.80, NNTH 8, CI 6 to 15) but not in long-term studies (2 RCTs, n = 646, RR 1.27 CI 0.97 to 1.66, Analysis 1.40). Heterogeneity might be due to the differences in type of typical antipsychotics.

1.41 Adverse effects: 7b. Weight gain - average weight in kg

There was no significant difference in short-term weight change (1 RCT, n = 25, MD 1.40 CI -5.66 to 8.46), long-term endpoint weight (1 RCT, n = 45, MD 4.30 CI -0.31 to 8.91) and long-term weight change (1 RCT, n = 98, MD 3.20 CI -1.79 to 8.19, Analysis 1.41).

1.42 Adverse effects: 8. Decreased white blood cell count - short term

There was no significant difference (1 RCT, n = 40, RR 0.33 CI 0.01 to 7.72, Analysis 1.42).

1.43 Sensitivity analysis

Excluding studies with skewed data (Analysis 1.43; Analysis 1.44), inappropriate comparator doses (Analysis 1.45; Analysis 1.46; Analysis 1.47), high attrition rate (Analysis 1.48; Analysis 1.49; Analysis 1.50), or studies with intention-to-treat analysis from the evaluation of the PANSS total score, the PANSS positive subscore and the PANSS negative subscore (Analysis 1.51; Analysis 1.52; Analysis 1.53) did not reveal markedly different results.  

Discussion

Summary of main results

The overall search strategy yielded 830 reports of which 65 were closely inspected. Forty-three studies with 7217 participants met the inclusion criteria. Twelve studies were sponsored by pharmaceutical companies developing quetiapine. The comparator typical antipsychotics were chlorpromazine, haloperidol and perphenazine.

1. Global state

There was no significant difference between groups on 'no clinical improvement'. The CGI-S and CGI-I scores suggested the superiority of typical antipsychotics in the short term. However, no significance difference of long-term CGI-S scores and medium-term CGI-I scores were found, suggesting that the two groups might not differ in terms of global improvement.

2. Leaving the study early

The overall attrition rates from any reason were high in both groups (36%), although they were similar to those found in previous analyses (Liu-Seifert 2005; Srisurapanont 2004). The high dropout rate may represent poor psychiatric response (Liu-Seifert 2005). This study found no significant difference between groups on leaving the study early due to inefficacy. However, quetiapine had a significantly lower risk of leaving the study early due to adverse events. This may indicate that quetiapine is more tolerable than typical antipsychotics.

3. Mental state

3.1 Positive symptoms

Overall, the positive symptoms in both groups were not different in short-term, medium-term and long-term treatment. Only one study (vs CPZ - AstraZeneca 2005) that used chlorpromazine as comparator found a significant difference in favour of typical antipsychotics. However, the point difference was about one, and the clinical relevance of this is unclear.

3.2 Negative symptoms

The data on negative symptoms were relatively heterogenous. The PANSS negative subscores of two studies (n = 88) (vs CPZ - Li 2003; vs HLP - Purdon 2001) favoured quetiapine in medium-term studies, but the analyses in short-term studies (19 studies, n = 1621) and one long-term study (one study, n = 45) found no significant difference.  BPRS negative subscores favoured quetiapine in one study (n = 253) (vs CPZ - AstraZeneca 2005) but not in the other (n = 25) (vs FLUPHEN - Conley 2005). In contrast, the results of a single short-term study (n = 103) using the SANS scale were in favour of haloperidol (vs HLP - Arvanitis 1997), while a chlorpromazine-controlled  study using the same scale failed to reveal any significant difference (n = 228) (vs CPZ - AstraZeneca 2005). The heterogeneity of these results might be due to small sample sizes and different types of typical antipsychotics. Overall, PANSS negative subscore (22 studies, n = 1934) and BPRS negative subscore (two studies, n = 278) showed significant differences in favour of quetiapine. However, the point difference was less than one, and the clinical relevance of this is unclear. Moreover, the PANSS negative subscore was highly heterogeneous and driven by two small outlier studies (vs CPZ - Jin 2007; vs CPZ - Zhang 2002). Without these two studies, there was no heterogeneity and no statistically significant difference between quetiapine and typical antipsychotics.

3.3 General psychopathology

PANSS general psychopathology subscore showed no significant difference between groups in general psychopathology for short-term, medium-term and long-term treatment.

4. General functioning and quality of life

Only limited data are available for these outcomes. GAF score and MANSA score in a long-term study (n = 207) (vs HLP - Fl'hacker 2005) as well as QLS score in another long-term study (n = 156) (vs PERPHEN - L'rman 2005) failed to reveal any significant difference between groups. However, the quality of life assessed by SF-36 in a single, short-term, chlorpromazine-controlled study (n = 84) (vs CPZ - Zhong 2005) was superior in the quetiapine group.

5. Cognitive function and service use

The results of a short-term, chlorpromazine-controlled study (n = 91) found the superiority of quetiapine in cognitive improvement (vs CPZ - Bai 2005); however, long-term outcomes from two studies (n = 51) were not significantly different between groups.

6. Adverse effects

Quetiapine was superior to typical antipsychotics for the measures of having at least one adverse effects, abnormal ECG, overall extrapyramidal symptoms, akathisia, parkinsonism, dystonia, tremor and prolactin level. However, no significant difference between groups was found on suicide attempt, suicide, death, QTc prolongation, low blood pressure, tachycardia, sedation, gynaecomastia, galactorrhoea and menstrual irregularity. Since some outcomes, i.e. suicide attempt, suicide and death, are rare events and measured only in a small number of studies, a significant difference between groups might not been revealed. This review found that quetiapine caused less weight gain in short-term studies (10 studies, n = 955) but not in long-term studies (two studies, n = 646). This might have been caused by the fact that chlorpromazine was the control in most short-term studies while haloperidol and fluphenazine were the controls for long-term studies.

Overall completeness and applicability of evidence

1. Completeness

We could identify only one low-powered study for some important outcomes; thus, the evidence is incomplete. The high attrition rates in many of the studies suggest the incompleteness of data. While schizophrenia is a life-long disorder, only three long-term studies have been carried out.

The emphasis on continuous measures leaves several questions unanswered. We remain unsure if quetiapine improves mental state, functioning, quality of life to any important extent.

2. Applicability

The applicability of this evidence is limited. Most of the studies were highly controlled explanatory trials, which may be of limit for the application of these results in the real world (Thorpe 2009).

Quality of the evidence

All studies were randomised, but their details were rarely presented. Only 10 studies were double-blinded, and it is unclear in almost all studies whether the randomisation and blinding were appropriately carried out. Eight studies had high risk of bias in terms of incomplete data. Twelve studies were sponsored by pharmaceutical companies. All these factors may well considerably limit the quality of evidence and this is reflected in our grading within Summary of findings for the main comparison.

Potential biases in the review process

It is perfectly feasible that we have failed to identify small relevant trials for many reasons (Easterbrook 1991), and we have not adjusted for this potential (Macaskill 2001). We feel that it is unlikely that large studies that would make a substantive difference to the results have been omitted. We could have been biased by our foreknowledge of data in Srisurapanont 2004 but, again, think that this potential for bias would have substantial effects on our extracting and writing of results.

Agreements and disagreements with other studies or reviews

A previous Cochrane review compared the efficacy between quetiapine and placebo, typical and atypical antipsychotic medications for schizophrenia (Srisurapanont 2004). The authors concluded that quetiapine was not much different from typical antipsychotics with respect to treatment withdrawal and efficacy. This review has included many recent studies. It is, therefore, more up-to date and more comprehensive. In Srisurapanont 2004, quetiapine was superior to typical antipsychotics only in respect of antipsychotic-induced movement disorders. In this review, we found the superiority of quetiapine on some more respects, e.g. having less overall adverse effects, abnormal ECG, prolactin level and weight gain (short term).

Authors' conclusions

Implications for practice

1. For people with schizophrenia

For people with schizophrenia it is important to know that quetiapine may be less likely to cause adverse effects, such as abnormal ECG, extrapyramidal effects, abnormal prolactin levels and weight gain (short term). However, its efficacy in treating positive symptoms and general psychopathology seems to be similar to that of typical antipsychotics. There is no clear differences in terms of the treatment of negative symptoms.

2. For clinicians

Clinicians should be aware that the evidence is limited due to high attrition rates in almost all studies. Well-designed studies comparing quetiapine with typical antipsychotics are needed. See Table 1 for a suggested study design.

Table 1. Suggested design of future study
  1. * Power calculation suggested 300/group would allow good chance of showing a 10% difference between groups for primary outcome.

    ** Of the many possible comparisons we would probably choose chlorpromazine or perphenazine.

    *** Primary outcome.
    CGI: Clinical Global Impression Scale
    PANSS: Positive and Negative Syndrome Scale

MethodsAllocation: randomised - clearly described generation of sequence and concealment of allocation.
Blindness.: double - described and tested.
Duration: 6 months minimum.
ParticipantsDiagnosis: schizophrenia.
N = 1800.*
Age: any.
Sex: both.
History: any.
Interventions

1. Quetiapine: dose 300-800 mg/day. N = 300.

2. Typical antipsychotic medications.

2a. Chlorpromazine: dose 300-1000 mg/day. N = 300.**

2b. Fluphenazine: dose 5-20 mg/day.N = 300.

2c. Perphenazine: dose 16-64 mg/day. N = 300.

2d. Trifluoperazine: dose 15-50 mg/day. N = 300.

2e. Haloperidol: dose 5-20 mg/day. N = 300.

Outcomes

Global impression: CGI***, relapse.

Leaving study early (any reason, adverse events, inefficacy).
Service outcomes: hospitalised, time in hospital, attending out patient clinics.
Mental state: PANSS.
Adverse events.

Functioning: employment, living independently, functioning improved to an important extent.

Quality of life: improved to an important extent.

Economics: direct and indirect costs.

3. For managers/policy makers

Limited data suggest no difference between quetiapine and typical antipsychotics in term of re-hospitalisation. This finding is important because in many countries hospitalisation is a main cost of schizophrenia treatment.

Implications for research

1. General

Outcome reporting remains insufficient in antipsychotic drug trials. Strict adherence to the CONSORT statement (Moher 2001) would improve the conduct, and reporting of clinical trials. This statement has been available since 1996 and there seems little justification for authors or journal editors ignoring its proven value.

2. Specific

Pragmatic, real world, randomised controlled trials should be carried out to determine the effectiveness of quetiapine in standard clinical practice (Thorpe 2009). Studies of medium- and long-term risks, including mortality and cost-effectiveness, are a priority.

Acknowledgements

We thank Prof. Clive Adams, Prof. Stefan Leucht, Bethany York and Claire Irving for their support during the development of the protocol and full review. We thank Samantha Roberts for her assistance in developing the search strategy for this review, and for her help in conducting the electronic searches. We would also like to thank all members of the Cochrane Schizophrenia Group editorial base for their editorial assistance.

The Schizophrenia Group produce a standard template for the methodology of their reviews. We have used this template and adapted it.

Data and analyses

Download statistical data

Comparison 1. QUETIAPINE versus TYPICAL ANTIPSYCHOTICS
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Global state: 1. No clinical improvement151607Risk Ratio (IV, Random, 95% CI)0.96 [0.75, 1.23]
1.1 short term151479Risk Ratio (IV, Random, 95% CI)1.05 [0.81, 1.35]
1.2 medium term1128Risk Ratio (IV, Random, 95% CI)0.16 [0.05, 0.51]
2 Global state: 2a. CGI: Average CGI-S (high = poor)81253Mean Difference (IV, Random, 95% CI)0.20 [0.06, 0.33]
2.1 short term - endpoint score5347Mean Difference (IV, Random, 95% CI)0.22 [-0.04, 0.49]
2.2 short term - change score2699Mean Difference (IV, Random, 95% CI)0.20 [0.04, 0.36]
2.3 long term - endpoint score1207Mean Difference (IV, Random, 95% CI)-0.10 [-0.93, 0.73]
3 Global state: 2b. CGI: Average endpoint CGI-I (high = poor)5507Mean Difference (IV, Random, 95% CI)0.38 [0.28, 0.47]
3.1 short term4496Mean Difference (IV, Random, 95% CI)0.39 [0.30, 0.47]
3.2 medium term111Mean Difference (IV, Random, 95% CI)-0.20 [-1.08, 0.68]
4 Leaving the study early: 1. Any reason233576Risk Ratio (IV, Random, 95% CI)0.91 [0.81, 1.01]
4.1 short term172535Risk Ratio (IV, Random, 95% CI)0.88 [0.78, 1.00]
4.2 medium term3203Risk Ratio (IV, Random, 95% CI)0.72 [0.52, 0.99]
4.3 long term3838Risk Ratio (IV, Random, 95% CI)0.99 [0.77, 1.28]
5 Leaving the study early: 2. Adverse events153010Risk Ratio (IV, Random, 95% CI)0.48 [0.30, 0.77]
5.1 short term112044Risk Ratio (IV, Random, 95% CI)0.51 [0.31, 0.84]
5.2 medium term1128Risk Ratio (IV, Random, 95% CI)0.03 [0.00, 0.53]
5.3 long term3838Risk Ratio (IV, Random, 95% CI)0.46 [0.12, 1.78]
6 Leaving the study early: 3. Inefficacy92063Risk Ratio (IV, Random, 95% CI)1.24 [1.00, 1.54]
6.1 short term3842Risk Ratio (IV, Random, 95% CI)1.20 [0.87, 1.66]
6.2 long term61221Risk Ratio (IV, Random, 95% CI)1.41 [0.94, 2.11]
7 Mental state: 1a. General - average score (PANSS total, high = poor)273052Mean Difference (IV, Random, 95% CI)0.09 [-2.14, 2.31]
7.1 short term - endpoint score212055Mean Difference (IV, Random, 95% CI)0.84 [-1.46, 3.15]
7.2 short term - change score2565Mean Difference (IV, Random, 95% CI)3.59 [0.09, 7.10]
7.3 medium term - endpoint score3180Mean Difference (IV, Random, 95% CI)-7.77 [-15.69, 0.14]
7.4 long term - endpoint score2252Mean Difference (IV, Random, 95% CI)-0.90 [-4.94, 3.15]
8 Mental state: 1b. General - average score -short term (BPRS total, high = poor)81025Mean Difference (IV, Random, 95% CI)0.71 [-0.77, 2.20]
8.1 endpoint score6666Mean Difference (IV, Random, 95% CI)-0.24 [-1.66, 1.17]
8.2 change score2359Mean Difference (IV, Random, 95% CI)3.49 [0.90, 6.08]
9 Mental state: 2a. Positive symptoms - average endpoint score (PANSS positive subscore, high = poor)221934Mean Difference (IV, Random, 95% CI)0.02 [-0.39, 0.43]
9.1 short term191801Mean Difference (IV, Random, 95% CI)0.19 [-0.25, 0.62]
9.2 medium term288Mean Difference (IV, Random, 95% CI)-1.57 [-3.41, 0.26]
9.3 long term145Mean Difference (IV, Random, 95% CI)-1.30 [-3.12, 0.52]
10 Mental state: 2b. Positive symptoms - average score - short term (BPRS positive subscore, high = poor)3381Mean Difference (IV, Random, 95% CI)0.63 [-0.37, 1.62]
10.1 endpoint score2128Mean Difference (IV, Random, 95% CI)0.12 [-0.42, 0.65]
10.2 change score1253Mean Difference (IV, Random, 95% CI)1.34 [0.27, 2.41]
11 Mental state: 3a. Negative symptoms - average endpoint score (PANSS negative subscore, high = poor)221934Mean Difference (IV, Random, 95% CI)-0.82 [-1.59, -0.04]
11.1 short term191801Mean Difference (IV, Random, 95% CI)-0.81 [-1.63, 0.01]
11.2 medium term288Mean Difference (IV, Random, 95% CI)-2.45 [-4.64, -0.27]
11.3 long term145Mean Difference (IV, Random, 95% CI)1.20 [-2.09, 4.49]
12 Mental state: 3b. Negative symptoms - average score - short term (BPRS negative subscore, high = poor))2278Mean Difference (IV, Random, 95% CI)-0.67 [-1.31, -0.04]
12.1 endpoint score125Mean Difference (IV, Random, 95% CI)0.11 [-2.01, 2.23]
12.2 change score1253Mean Difference (IV, Random, 95% CI)-0.75 [-1.42, -0.08]
13 Mental state: 3c. Negative symptoms - average score - short term (SANS total, high = poor)2331Mean Difference (IV, Random, 95% CI)0.66 [-1.35, 2.66]
13.1 endpoint score1103Mean Difference (IV, Random, 95% CI)1.80 [0.24, 3.36]
13.2 change score1228Mean Difference (IV, Random, 95% CI)-0.26 [-1.08, 0.56]
14 Mental state: 4. General psychopathology - average endpoint score (PANSS general psychopathology subscore, high = poor)181569Mean Difference (IV, Random, 95% CI)-0.20 [-0.83, 0.42]
14.1 short term151472Mean Difference (IV, Random, 95% CI)-0.17 [-0.81, 0.47]
14.2 medium term252Mean Difference (IV, Random, 95% CI)0.47 [-3.30, 4.24]
14.3 long term145Mean Difference (IV, Random, 95% CI)-2.20 [-6.02, 1.62]
15 General functioning: General - average endpoint score - long term (GAF total score, low = poor)1207Mean Difference (IV, Random, 95% CI)-0.10 [-9.80, 9.60]
16 Quality of life: 1a. General - average endpoint score -short term (SF-36, low = poor)184Mean Difference (IV, Random, 95% CI)13.76 [5.95, 21.57]
17 Quality of life: 1b. General - average endpoint score - long term (MANSA total score, low = poor)1207Mean Difference (IV, Random, 95% CI)0.0 [-1.38, 1.38]
18 Quality of life: 1c. General - average change score - long term (QLS, high = poor)1156Mean Difference (IV, Random, 95% CI)0.1 [-0.23, 0.43]
19 Cognitive function: 1a. General - average change score - long term (Composite score, high = poor)2407Mean Difference (IV, Random, 95% CI)0.00 [-0.19, 0.20]
20 Cognitive function: 1b. General - average endpoint scores as defined by the original studies (low = poor)3142Mean Difference (IV, Random, 95% CI)1.55 [-0.62, 3.72]
20.1 short term191Mean Difference (IV, Random, 95% CI)8.30 [1.64, 14.96]
20.2 long term251Mean Difference (IV, Random, 95% CI)0.82 [-0.77, 2.41]
21 Service use: number of participants re-hospitalised - long term2722Risk Ratio (IV, Random, 95% CI)1.23 [0.90, 1.68]
22 Adverse effects: 1. General - at least one adverse effect91985Risk Ratio (IV, Random, 95% CI)0.76 [0.64, 0.90]
22.1 short term71180Risk Ratio (IV, Random, 95% CI)0.75 [0.62, 0.91]
22.2 long term2805Risk Ratio (IV, Random, 95% CI)0.82 [0.52, 1.29]
23 Adverse effects: 2. Death21456Risk Ratio (IV, Random, 95% CI)0.78 [0.20, 3.04]
23.1 suicide attempt - long term1598Risk Ratio (IV, Random, 95% CI)0.77 [0.05, 12.32]
23.2 suicide - long term1598Risk Ratio (IV, Random, 95% CI)0.52 [0.09, 3.07]
23.3 death - short term1260Risk Ratio (IV, Random, 95% CI)3.0 [0.12, 72.97]
24 Adverse effects: 3a. Cardiac effects - QTc prolongation - long term141Risk Ratio (IV, Random, 95% CI)1.73 [0.17, 17.59]
25 Adverse effects: 3b. Cardiac effects - abnormal ECG - short term2165Risk Ratio (IV, Random, 95% CI)0.38 [0.16, 0.92]
26 Adverse effects: 3c. Cardiac effects - orthostatic hypotension51393Risk Ratio (IV, Random, 95% CI)0.61 [0.28, 1.35]
26.1 short term4795Risk Ratio (IV, Random, 95% CI)0.45 [0.13, 1.54]
26.2 long term1598Risk Ratio (IV, Random, 95% CI)1.01 [0.64, 1.60]
27 Adverse effects: 3d. Cardiac effects - low blood pressure - short term6572Risk Ratio (IV, Random, 95% CI)0.69 [0.41, 1.18]
28 Adverse effects: 3e. Cardiac effects - tachycardia - short term8814Risk Ratio (IV, Random, 95% CI)0.70 [0.36, 1.34]
29 Adverse effects: 4. Central nervous system - sedation142298Risk Ratio (IV, Random, 95% CI)0.70 [0.43, 1.16]
29.1 short term131700Risk Ratio (IV, Random, 95% CI)0.69 [0.37, 1.30]
29.2 long term1598Risk Ratio (IV, Random, 95% CI)1.08 [0.84, 1.39]
30 Adverse effects: 5a. Extrapyramidal effects - overall - short term81095Risk Ratio (IV, Random, 95% CI)0.17 [0.09, 0.32]
31 Adverse effects: 5b. Extrapyramidal effects - akathisia162217Risk Ratio (IV, Random, 95% CI)0.27 [0.18, 0.39]
31.1 short term152059Risk Ratio (IV, Random, 95% CI)0.24 [0.17, 0.35]
31.2 long term1158Risk Ratio (IV, Random, 95% CI)0.50 [0.25, 0.98]
32 Adverse effects: 5c. Extrapyramidal effects - parkinsonism3343Risk Ratio (IV, Random, 95% CI)0.26 [0.14, 0.47]
32.1 short term2185Risk Ratio (IV, Random, 95% CI)0.13 [0.04, 0.48]
32.2 long term1158Risk Ratio (IV, Random, 95% CI)0.31 [0.15, 0.62]
33 Adverse effects: 5d. Extrapyramidal effects - dystonia5992Risk Ratio (IV, Random, 95% CI)0.19 [0.06, 0.64]
33.1 short term4834Risk Ratio (IV, Random, 95% CI)0.13 [0.04, 0.51]
33.2 long term1158Risk Ratio (IV, Random, 95% CI)0.86 [0.05, 13.49]
34 Adverse effects: 5e. Extrapyramidal effects - tremor - short term121641Risk Ratio (IV, Random, 95% CI)0.33 [0.23, 0.47]
35 Adverse effects: 5f. Extrapyramidal effects - scale measured (dichotomous data) - long term1 Risk Ratio (IV, Random, 95% CI)Subtotals only
35.1 extrapyramidal symptoms: Simpson-Angus Scale (SAS >/=1)1541Risk Ratio (IV, Random, 95% CI)0.65 [0.31, 1.37]
35.2 abnormal movement: Abnormal Involuntary Movement Scale (AIMS>/=2)1473Risk Ratio (IV, Random, 95% CI)0.73 [0.48, 1.14]
35.3 akathisia: Barnes Akathisia Scale (BARS >/=3)1546Risk Ratio (IV, Random, 95% CI)0.79 [0.40, 1.55]
36 Adverse effects: 5g. Extrapyramidal effects - scale measured (continuous data, high=poor)5373Mean Difference (IV, Random, 95% CI)-1.24 [-2.54, 0.05]
36.1 average endpoint score - Extrapyramidal symptoms (ESRS) - short term135Mean Difference (IV, Random, 95% CI)-4.55 [-6.58, -2.52]
36.2 average endpoint score - Abnormal Involuntary Movement Scale ( AIMS) - medium term111Mean Difference (IV, Random, 95% CI)0.30 [-0.83, 1.43]
36.3 average endpoint score - Simpson-Angus Scale (SAS) - medium term111Mean Difference (IV, Random, 95% CI)1.10 [-1.94, 4.14]
36.4 average endpoint score - Treatment Emergent Symptom Scale (TESS) - short term2158Mean Difference (IV, Random, 95% CI)-1.09 [-2.51, 0.33]
36.5 average change score - Treatment Emergent Symptom Scale (TESS) - short term1117Mean Difference (IV, Random, 95% CI)-1.03 [-2.49, 0.43]
36.6 average endpoint score - Treatment Emergent Symptom Scale (TESS) - medium term141Mean Difference (IV, Random, 95% CI)-2.1 [-4.55, 0.35]
37 Adverse effects: 6a. Prolactin associated side effects11196Risk Ratio (IV, Random, 95% CI)0.77 [0.33, 1.79]
37.1 gynaecomastia, galactorrhoea - long term1598Risk Ratio (IV, Random, 95% CI)1.16 [0.33, 4.07]
37.2 menstrual irregularities - long term1598Risk Ratio (IV, Random, 95% CI)0.55 [0.18, 1.72]
38 Adverse effects: 6b. Prolactin - Hyperprolactinemia3229Risk Ratio (IV, Random, 95% CI)0.27 [0.04, 1.92]
38.1 short term2165Risk Ratio (IV, Random, 95% CI)0.08 [0.01, 0.60]
38.2 long term164Risk Ratio (IV, Random, 95% CI)0.91 [0.51, 1.62]
39 Adverse effects: 6c. Prolactin level - average level in ng/mL41034Mean Difference (IV, Random, 95% CI)-16.20 [-23.34, -9.07]
39.1 short term - endpoint level135Mean Difference (IV, Random, 95% CI)-15.67 [-21.00, -10.34]
39.2 short term - change level1356Mean Difference (IV, Random, 95% CI)-22.43 [-23.20, -21.66]
39.3 long term - endpoint level145Mean Difference (IV, Random, 95% CI)-16.4 [-23.83, -8.97]
39.4 long term - change level1598Mean Difference (IV, Random, 95% CI)-9.70 [-14.02, -5.38]
40 Adverse effects: 7a. Weight gain (as defined by the original studies)111512Risk Ratio (IV, Random, 95% CI)0.68 [0.44, 1.04]
40.1 short term9866Risk Ratio (IV, Random, 95% CI)0.52 [0.34, 0.80]
40.2 long term2646Risk Ratio (IV, Random, 95% CI)1.27 [0.97, 1.66]
41 Adverse effects: 7b. Weight gain - average weight in kg.3168Mean Difference (IV, Random, 95% CI)3.35 [0.29, 6.40]
41.1 short term - change weight125Mean Difference (IV, Random, 95% CI)1.40 [-5.66, 8.46]
41.2 long term - endpoint weight145Mean Difference (IV, Random, 95% CI)4.30 [-0.31, 8.91]
41.3 long term - change weight198Mean Difference (IV, Random, 95% CI)3.2 [-1.79, 8.19]
42 Adverse effects: 8. decreased white blood cell count - short term140Risk Ratio (IV, Random, 95% CI)0.33 [0.01, 7.72]
43 Sensitivity analysis (skewed data excluded), Mental state: 1. General - average endpoint score (PANSS total, high = poor)3172Mean Difference (IV, Random, 95% CI)-2.25 [-5.35, 0.85]
43.1 short term295Mean Difference (IV, Random, 95% CI)-0.88 [-3.63, 1.88]
43.2 medium term177Mean Difference (IV, Random, 95% CI)-4.90 [-8.06, -1.74]
44 Sensitivity analysis (skewed data excluded), Mental state: 2. Negative symptoms - average endpoint score (PANSS negative subscore, high=poor) - medium term177Mean Difference (IV, Random, 95% CI)-2.70 [-4.96, -0.44]
45 Sensitivity analysis (inappropriate comparator doses excluded), Mental state: 1: General- average endpoint score (PANSS total, high=poor)171835Mean Difference (IV, Random, 95% CI)1.55 [-1.38, 4.49]
45.1 short term - endpoint131173Mean Difference (IV, Random, 95% CI)2.04 [-1.67, 5.74]
45.2 short term - change2565Mean Difference (IV, Random, 95% CI)3.59 [0.09, 7.10]
45.3 medium term - endpoint252Mean Difference (IV, Random, 95% CI)-4.72 [-7.85, -1.59]
45.4 long term - endpoint145Mean Difference (IV, Random, 95% CI)-2.30 [-10.22, 5.62]
46 Sensitivity analysis (inappropriate comparator doses excluded), Mental state: 2: Positive symptoms- average endpoint score (PANSS positive subscore, high = poor)131011Mean Difference (IV, Random, 95% CI)0.02 [-0.58, 0.62]
47 Sensitivity analysis (inappropriate comparator doses excluded), Mental state: 3: Negative symptoms- average endpoint score (PANSS negative subscore, high = poor)131011Mean Difference (IV, Random, 95% CI)-0.98 [-2.09, 0.14]
47.1 short term10878Mean Difference (IV, Random, 95% CI)-0.99 [-2.26, 0.28]
47.2 medium term288Mean Difference (IV, Random, 95% CI)-2.45 [-4.64, -0.27]
47.3 long term145Mean Difference (IV, Random, 95% CI)1.20 [-2.09, 4.49]
48 Senstivity analysis (high attrition rate data excluded), Mental state: 1: General - average score (PANSS total, high = poor)252834Mean Difference (IV, Random, 95% CI)0.08 [-2.26, 2.42]
48.1 short term - endpoint score212055Mean Difference (IV, Random, 95% CI)0.84 [-1.46, 3.15]
48.2 short term - change score2565Mean Difference (IV, Random, 95% CI)3.59 [0.09, 7.10]
48.3 medium term - endpoint score2169Mean Difference (IV, Random, 95% CI)-9.11 [-17.54, -0.68]
48.4 long term - endpoint score145Mean Difference (IV, Random, 95% CI)-2.30 [-10.22, 5.62]
49 Sensitivity analysis (high attrition rate data excluded), Mental state: 2. Positive symptoms - average endpoint score (PANSS positive subscore, high = poor)211923Mean Difference (IV, Random, 95% CI)0.02 [-0.39, 0.43]
49.1 short term191801Mean Difference (IV, Random, 95% CI)0.19 [-0.25, 0.62]
49.2 medium term177Mean Difference (IV, Random, 95% CI)-1.80 [-3.70, 0.10]
49.3 long term145Mean Difference (IV, Random, 95% CI)-1.30 [-3.12, 0.52]
50 Sensitivity analysis (high attrition rate data excluded), Mental state: 3. Negative symptoms - average endpoint score (PANSS negative subscore, high = poor)211923Mean Difference (IV, Random, 95% CI)-0.83 [-1.62, -0.04]
50.1 short term191801Mean Difference (IV, Random, 95% CI)-0.81 [-1.63, 0.01]
50.2 medium term177Mean Difference (IV, Random, 95% CI)-2.70 [-4.96, -0.44]
50.3 long term145Mean Difference (IV, Random, 95% CI)1.20 [-2.09, 4.49]
51 Senstivity analysis (completer data only), Mental state: 1: General - average score (PANSS total, high=poor)181628Mean Difference (IV, Random, 95% CI)-0.41 [-3.32, 2.51]
51.1 short term - endpoint score161342Mean Difference (IV, Random, 95% CI)0.55 [-2.27, 3.38]
51.2 short term - change score1117Mean Difference (IV, Random, 95% CI)3.88 [-1.68, 9.44]
51.3 medium term - endpoint score2169Mean Difference (IV, Random, 95% CI)-9.11 [-17.54, -0.68]
52 Sensitivity analysis (completer data only), Mental state: 2. Positive symptoms - average endpoint score (PANSS positive subscore, high = poor)161385Mean Difference (IV, Random, 95% CI)0.00 [-0.46, 0.46]
52.1 short term151308Mean Difference (IV, Random, 95% CI)0.11 [-0.36, 0.59]
52.2 medium term177Mean Difference (IV, Random, 95% CI)-1.80 [-3.70, 0.10]
53 Sensitivity analysis (completer data only), Mental state: 3. Negative symptoms - average endpoint score (PANSS negative subscore, high = poor)161385Mean Difference (IV, Random, 95% CI)-1.16 [-2.08, -0.25]
53.1 short term151308Mean Difference (IV, Random, 95% CI)-1.06 [-2.01, -0.12]
53.2 medium term177Mean Difference (IV, Random, 95% CI)-2.70 [-4.96, -0.44]
Analysis 1.1.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 1 Global state: 1. No clinical improvement.

Analysis 1.2.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 2 Global state: 2a. CGI: Average CGI-S (high = poor).

Analysis 1.3.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 3 Global state: 2b. CGI: Average endpoint CGI-I (high = poor).

Analysis 1.4.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 4 Leaving the study early: 1. Any reason.

Analysis 1.5.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 5 Leaving the study early: 2. Adverse events.

Analysis 1.6.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 6 Leaving the study early: 3. Inefficacy.

Analysis 1.7.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 7 Mental state: 1a. General - average score (PANSS total, high = poor).

Analysis 1.8.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 8 Mental state: 1b. General - average score -short term (BPRS total, high = poor).

Analysis 1.9.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 9 Mental state: 2a. Positive symptoms - average endpoint score (PANSS positive subscore, high = poor).

Analysis 1.10.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 10 Mental state: 2b. Positive symptoms - average score - short term (BPRS positive subscore, high = poor).

Analysis 1.11.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 11 Mental state: 3a. Negative symptoms - average endpoint score (PANSS negative subscore, high = poor).

Analysis 1.12.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 12 Mental state: 3b. Negative symptoms - average score - short term (BPRS negative subscore, high = poor)).

Analysis 1.13.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 13 Mental state: 3c. Negative symptoms - average score - short term (SANS total, high = poor).

Analysis 1.14.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 14 Mental state: 4. General psychopathology - average endpoint score (PANSS general psychopathology subscore, high = poor).

Analysis 1.15.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 15 General functioning: General - average endpoint score - long term (GAF total score, low = poor).

Analysis 1.16.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 16 Quality of life: 1a. General - average endpoint score -short term (SF-36, low = poor).

Analysis 1.17.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 17 Quality of life: 1b. General - average endpoint score - long term (MANSA total score, low = poor).

Analysis 1.18.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 18 Quality of life: 1c. General - average change score - long term (QLS, high = poor).

Analysis 1.19.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 19 Cognitive function: 1a. General - average change score - long term (Composite score, high = poor).

Analysis 1.20.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 20 Cognitive function: 1b. General - average endpoint scores as defined by the original studies (low = poor).

Analysis 1.21.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 21 Service use: number of participants re-hospitalised - long term.

Analysis 1.22.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 22 Adverse effects: 1. General - at least one adverse effect.

Analysis 1.23.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 23 Adverse effects: 2. Death.

Analysis 1.24.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 24 Adverse effects: 3a. Cardiac effects - QTc prolongation - long term.

Analysis 1.25.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 25 Adverse effects: 3b. Cardiac effects - abnormal ECG - short term.

Analysis 1.26.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 26 Adverse effects: 3c. Cardiac effects - orthostatic hypotension.

Analysis 1.27.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 27 Adverse effects: 3d. Cardiac effects - low blood pressure - short term.

Analysis 1.28.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 28 Adverse effects: 3e. Cardiac effects - tachycardia - short term.

Analysis 1.29.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 29 Adverse effects: 4. Central nervous system - sedation.

Analysis 1.30.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 30 Adverse effects: 5a. Extrapyramidal effects - overall - short term.

Analysis 1.31.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 31 Adverse effects: 5b. Extrapyramidal effects - akathisia.

Analysis 1.32.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 32 Adverse effects: 5c. Extrapyramidal effects - parkinsonism.

Analysis 1.33.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 33 Adverse effects: 5d. Extrapyramidal effects - dystonia.

Analysis 1.34.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 34 Adverse effects: 5e. Extrapyramidal effects - tremor - short term.

Analysis 1.35.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 35 Adverse effects: 5f. Extrapyramidal effects - scale measured (dichotomous data) - long term.

Analysis 1.36.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 36 Adverse effects: 5g. Extrapyramidal effects - scale measured (continuous data, high=poor).

Analysis 1.37.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 37 Adverse effects: 6a. Prolactin associated side effects.

Analysis 1.38.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 38 Adverse effects: 6b. Prolactin - Hyperprolactinemia.

Analysis 1.39.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 39 Adverse effects: 6c. Prolactin level - average level in ng/mL.

Analysis 1.40.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 40 Adverse effects: 7a. Weight gain (as defined by the original studies).

Analysis 1.41.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 41 Adverse effects: 7b. Weight gain - average weight in kg..

Analysis 1.42.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 42 Adverse effects: 8. decreased white blood cell count - short term.

Analysis 1.43.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 43 Sensitivity analysis (skewed data excluded), Mental state: 1. General - average endpoint score (PANSS total, high = poor).

Analysis 1.44.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 44 Sensitivity analysis (skewed data excluded), Mental state: 2. Negative symptoms - average endpoint score (PANSS negative subscore, high=poor) - medium term.

Analysis 1.45.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 45 Sensitivity analysis (inappropriate comparator doses excluded), Mental state: 1: General- average endpoint score (PANSS total, high=poor).

Analysis 1.46.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 46 Sensitivity analysis (inappropriate comparator doses excluded), Mental state: 2: Positive symptoms- average endpoint score (PANSS positive subscore, high = poor).

Analysis 1.47.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 47 Sensitivity analysis (inappropriate comparator doses excluded), Mental state: 3: Negative symptoms- average endpoint score (PANSS negative subscore, high = poor).

Analysis 1.48.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 48 Senstivity analysis (high attrition rate data excluded), Mental state: 1: General - average score (PANSS total, high = poor).

Analysis 1.49.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 49 Sensitivity analysis (high attrition rate data excluded), Mental state: 2. Positive symptoms - average endpoint score (PANSS positive subscore, high = poor).

Analysis 1.50.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 50 Sensitivity analysis (high attrition rate data excluded), Mental state: 3. Negative symptoms - average endpoint score (PANSS negative subscore, high = poor).

Analysis 1.51.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 51 Senstivity analysis (completer data only), Mental state: 1: General - average score (PANSS total, high=poor).

Analysis 1.52.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 52 Sensitivity analysis (completer data only), Mental state: 2. Positive symptoms - average endpoint score (PANSS positive subscore, high = poor).

Analysis 1.53.

Comparison 1 QUETIAPINE versus TYPICAL ANTIPSYCHOTICS, Outcome 53 Sensitivity analysis (completer data only), Mental state: 3. Negative symptoms - average endpoint score (PANSS negative subscore, high = poor).

Appendices

Appendix 1. Data collection and analysis published in protocol

Data collection and analysis  

Selection of studies  

Two review authors will independently inspect citations identified from the search. We will identify potentially relevant reports and order full papers for reassessment. Retrieved articles will be independently assessed by the two review authors for inclusion according to our defined inclusion criteria. Any disagreement will be resolved by consensus discussions with the third member of the review team. If it is impossible to resolve disagreements, these studies will be added to those awaiting assessment and we will contact the authors of the papers for clarification. Non-concurrence in trial selection will be reported.

Data extraction and management  

1. Extraction
Two review authors will independently extract data from included studies. Again, any disagreement will be discussed with the third member of the review team, decisions documented and, if necessary, we will contact authors of studies for clarification.

2. Management
Data will be extracted onto standard, simple forms.

3. Scale-derived data
We will include continuous data from rating scales only if: (a) the psychometric properties of the measuring instrument has been described in a peer-reviewed journal (Marshall 2000); (b) the measuring instrument was not written or modified by one of the trialists; (c) the measuring instrument was either (i) a self-report or (ii) completed by an independent rater or relative (not the therapist).

Assessment of risk of bias in included studies  

Two review authors will independently assess risk of bias in accordance with The Cochrane Collaboration’s tool for assessing quality and risk of bias (Higgins 2008). This tool encourages consideration of how the sequence was generated, how allocation was concealed, the integrity of blinding, the completeness of outcome data, selective reporting and other biases.

The risk of bias in each domain and overall will be assessed and categorised into:

A. low risk of bias: plausible bias unlikely to seriously alter the results (categorised as 'Yes' in 'Risk of bias' table);
B. high risk of bias: plausible bias that seriously weakens confidence in the results (categorised as 'No' in 'Risk of bias' table);
C. unclear risk of bias: plausible bias that raises some doubt about the results (categorised as 'Unclear' in 'Risk of bias' table).

Trials with a high risk of bias (defined as at least three out of five domains categorised as 'No') will not be included in the meta-analysis. If the raters disagree, the final rating will be made by consensus with the involvement of another member of the review group. Where inadequate details of randomisation and other characteristics of trials are provided, we will contact authors of the studies in order to obtain further information. Non-concurrence in quality assessment will be reported.

Measures of treatment effect  

1. Dichotomous data.
Where possible, efforts will be made to convert outcome measures to dichotomous data. This may be done by identifying cut-off points on rating scales and dividing participants accordingly into 'clinically improved' or 'not clinically improved'. It will be generally assumed that if there had been a 50% reduction in a scale-derived score such as the Brief Psychiatric Rating Scale (BPRS, Overall 1962) or the Positive and Negative Syndrome Scale (PANSS, Kay 1986), this could be considered as a clinically significant response (Leucht 2005a; Leucht 2005b). If data based on these thresholds are not available, we will use the primary cut-off presented by the original authors.

We will calculate the risk ratio (RR) and its 95% confidence interval (CI) based on the random-effects model, as this takes into account any differences between studies, even if there is no statistically significant heterogeneity. It has been shown that RR is more intuitive (Boissel 1999) than odds ratios and that odds ratios tend to be interpreted as RR by clinicians (Deeks 2000). This misinterpretation then leads to an overestimate of the impression of the effect. If the overall results are significant, we will calculate the number needed to treat to provide benefit (NNTB) and the number needed to treat to induce harm (NNTH) as the inverse of the risk difference

2. Continuous data.
2.1 Summary statistic
For continuous outcomes we will estimate a mean difference (MD) between groups. Mean differences will be based on the random-effects model as this takes into account any differences between studies even if there is no statistically significant heterogeneity. We will not calculate standardised mean differences measures.

2.2 Endpoint versus change data
Since there is no principal statistical reason why endpoint and change data should measure different effects (Higgins 2008), we will use scale endpoint data which is easier to interpret from a clinical point of view. If endpoint data are not available, we will use changed data.

2.3 Skewed data
Continuous data on clinical and social outcomes are often not normally distributed. To avoid the pitfall of applying parametric tests to non-parametric data, we will apply the following standards to all data before inclusion:
(a) standard deviations and means are reported in the paper or obtainable from the authors;
(b) when a scale starts from the finite number zero, the standard deviation, when multiplied by two, is less than the mean (as otherwise the mean is unlikely to be an appropriate measure of the centre of the distribution, (Altman 1996);
(c) if a scale starts from a positive value (such as PANSS which can have values from 30 to 210), the calculation described above will be modified to take the scale starting point into account. In these cases skew is present if 2,SD,>,(S-S min), where S is the mean score and S min is the minimum score.

Endpoint scores on scales often have a finite start and end point and these rules can be applied. When continuous data are presented on a scale that includes a possibility of negative values (such as change data), it is difficult to tell whether data are skewed or not. Skewed data from studies of less than 200 participants will be entered in additional tables rather than into an analysis. Skewed data pose less of a problem when looking at means if the sample size is large and will be entered into syntheses.

2.4 Data synthesis
If standard errors instead of standard deviations are presented, the former will be converted to standard deviations. If standard deviations are not reported and can not be calculated from available data, we will ask the study authors to supply the data. In the absence of data from authors, the mean standard deviations from other studies will be used.

2.5 Multiple doses
If a study has investigated a number of fixed doses of quetiapine, the scores from the highest dose group will be used.

Unit of analysis issues  

1. Cluster trials
Studies increasingly employ 'cluster randomisation' (such as randomisation by clinician or practice) but analysis and pooling of clustered data poses problems. Firstly, authors often fail to account for intraclass correlation in clustered studies, leading to a 'unit of analysis' error (Divine 1992) whereby P values are spuriously low, confidence intervals unduly narrow and statistical significance overestimated. This can lead to type I error or a false positive (Bland 1997; Gulliford 1999).

If clustering is not accounted for in primary studies, we will present the data in a table, with a (*) symbol to indicate the presence of a probable unit of analysis error. In subsequent versions of this review we will seek to contact first authors of studies to obtain intraclass correlation coefficients (ICCs) of their clustered data and will adjust for this using accepted method (Gulliford 1999). Where clustering has been incorporated into the analysis of primary studies, we will also present these data as if from a non-cluster randomised study, but will adjust for the clustering effect.

Binary data as presented in a report will be divided by a 'design effect'. This will be calculated using the mean number of participants per cluster (m) and the ICC [Design effect = 1+(m-1)*ICC] (Donner 2002). If the ICC is not reported, it will be assumed to be 0.1 (Ukoumunne 1999). This assumption may be too high and, should this instance occur, we will see if taking an ICC of 0.01 would make any substantive difference for the primary outcome. If it does, we will use 0.01 in preference across outcomes.

If cluster studies have been appropriately analysed taking into account ICCs and relevant data documented in the report, we will synthesise these with other studies using the generic inverse variance technique.

2. Cross-over trials
A major concern of cross-over trials is the carry-over effect. It occurs if an effect (e.g. pharmacological, physiological or psychological) of the treatment in the first phase is carried over to the second phase. As a consequence on entry to the second phase the participants can differ systematically from their initial state despite a wash-out phase. For the same reason cross-over trials are not appropriate if the condition of interest is unstable (Elbourne 2002). As both effects are very likely in schizophrenia, we will only use data of the first phase of cross-over studies.

3. Studies with multiple treatment groups
Where a study involves more than two treatment groups, if relevant, the additional treatment groups will be presented in additional relevant comparisons. Data will not be double counted. Where the additional treatment groups are not relevant, these data will not be reproduced.

Dealing with missing data  

1. Overall loss of credibility
At some degree of loss of follow-up, data must lose credibility (Xia 2007 - direct link). We are forced to make a judgment where this is for the short-term trials likely to be included in this review. Should more than 40% of data be unaccounted for, we willnot reproduce these data or use them within analyses.

2. Binary
In the case where attrition for a binary outcome is between 0% and 40% and outcomes of these people are described, we will include these data as reported. Where these data are not clearly described, we will present data on a 'once-randomised-always-analyse' basis, assuming an intention-to-treat (ITT) analysis. Those lost to follow-up will be assumed to have a negative outcome, with the exception of the outcome of death. For example, for the outcome of relapse, those lost to follow-up will all be considered as relapsed. A final sensitivity analysis will be undertaken to test how prone the primary outcomes are to change when 'completed' data only are compared with the ITT analysis using the negative assumption.

3. Continuous
In the case where attrition for a continuous outcome is between 0% and 40% and completer-only data are reported, we will reproduce these.

4. Intention-to-treat (ITT)
Intention-to-treat analysis will be used when available. We anticipate that in some studies, in order to do an ITT analysis, the method of last-observation-carried-forward (LOCF) will have been be employed within the study report. As with all methods of imputation to deal with missing data, LOCF introduces uncertainty about the reliability of the results. Therefore, if LOCF data are used in the analysis, we will indicate this in the review.

Assessment of heterogeneity  

1. Clinical heterogeneity
We will consider all included studies with the intention of using all studies together. Should clear unforeseen issues be apparent that may add obvious clinical heterogeneity, we will note these issues, consider them in analyses and undertake sensitivity analyses for the primary outcome.

2. Statistical
2.1 Visual inspection
We will visually inspect graphs to investigate the possibility of statistical heterogeneity.

2.2 Employing the I-squared (I2) statistic
Heterogeneity between studies will be investigated by considering the I2 method alongside the Chi2 'P' value. The I2 provides an estimate of the percentage of inconsistency thought to be due to chance (Higgins 2003). The importance of the observed value of I2 depends on i. magnitude and direction of effects and ii. strength of evidence for heterogeneity (e.g. 'P' value from Chi2 test, or a confidence interval for I2).

An I2 estimate greater than or equal to 50% accompanied by a statistically significant Chi2 statistic, will be interpreted as evidence of substantial levels of heterogeneity (Section 9.5.2 - Higgins 2008) and reasons for heterogeneity will be explored. If the inconsistency is high and the clear reasons are found, we will present data separately.

Assessment of reporting biases  

Reporting biases arise when the dissemination of research findings is influenced by the nature and direction of results (Egger 1997). These are described in section 10.1 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008). We are aware that funnel plots may be useful in investigating small-study effects but are of limited power to detect such effects when there are few studies. We will not use funnel plots for outcomes where there are 10 or fewer studies, or where all studies were of similar sizes. In other cases, where funnel plots are possible, we will seek statistical advice in their interpretation.

Data synthesis  

Where possible, we will employ a random-effects model for analyses. We understand that there is no closed argument for preference for use of fixed-effect or random-effects models. The random-effects method incorporates an assumption that the different studies are estimating different, yet related, intervention effects. This does seem true to us, however, random-effects does put added weight onto the smaller of the studies - those trials that are most vulnerable to bias.

Subgroup analysis and investigation of heterogeneity  

If data are clearly heterogeneous, we will check that data are correctly extracted and entered and that we have made no unit-of-analysis errors. If high levels of heterogeneity remain, we will not undertake a meta-analysis at this point for if there is considerable variation in results, and particularly if there is inconsistency in the direction of effect, it may be misleading to quote an average value for the intervention effect. However, we would want to explore heterogeneity. We have not pre-specified any characteristics of studies that may be associated with heterogeneity except quality of trial method. If no clear association can be shown by sorting studies by quality of methods, we will continue to investigate for other reasons for the heterogeneity. Should another characteristic of the studies be highlighted by the investigation of heterogeneity, perhaps some clinical heterogeneity not hitherto predicted - but plausible causes of heterogeneity - these post-hoc reasons will be discussed, the sensitivity of the estimate of effect size for the primary outcome to inclusion and exclusion of these causes investigated and data analysed and presented. However, should the heterogeneity be substantially unaffected by any investigation and no reasons for the heterogeneity being clear, the final data will be presented without a meta-analysis.

Sensitivity analysis  

We have planned sensitivity analyses a priori for examining the change in the robustness of the sensitivity to include studies with implied randomisation (see Criteria for considering studies for this review: Types of studies), skewed and non-skewed data, inappropriate comparator doses of drug and different clinical groups - the latter being defined post hoc.

If inclusion of studies with implied randomisation makes no substantive difference to the primary outcome, they will be left in the final analyses. For outcomes with both skewed data and non-skewed data, we will investigate the effect of combining all data together and if no substantive difference is noted then the potentially skewed data will be left in the analyses. A recent review showed that some of the comparisons of antipsychotics may have been biased by using inappropriate comparator dose ranges (Heres 2006). The inappropriate dose ranges were defined as the ranges not within the range recommended in the American Psychiatric Association Practice Guideline for the treatment of patients with Schizophrenia, second edition (APA 2004): 300-800 mg/day of quetiapine, 300-1000 mg/day of chlorpromazine, 5-20 mg/day of Fluphenazine, 150-400 mg/day of mesoridazine, 16-64 mg/day of perphenazine, 300-800 mg/day of thioridazine, 15-50 mg/day of trifluoperazine, 5-20 mg/day of haloperidol, 30-100 mg/day of loxapine, 30-100 mg/day of molindone and 15-50 mg/day of thiothixene. If we find studies with implied randomisation, skewed and non-skewed data, inappropriate comparator doses of drug and different clinical groups, we will analyse whether the exclusion of these studies changes the results of the primary outcome and the general mental state.

Contributions of authors

Sirijit Suttajit - protocol development, study selection, data extraction, analysis and writing of the final report.

Manit Srisurapanont - protocol development, study selection and writing of the final report.

Jun Xia - study selection, data extraction, analysis and writing of the final report.

Siritree Suttajit - protocol development, study selection and writing of the final report.

Benchalak Maneeton - protocol development, data extraction and writing of the final report.

Narong Maneeton - protocol development and writing of the final report.

Declarations of interest

Sirijit Suttajit and Manit Srisurapanont received support for attending national and international scientific meetings from AstraZeneca (Thailand), Eli Lilly Asia, Inc. (Thailand), GlaxoSmithKline (Thailand), Janssen-Cilag (Thailand) and Servier (Thailand).

Benchalak Maneeton and Narong Maneeton received similar supports for attending national scientific meetings from GlaxoSmithKline (Thailand) and Janssen-Cilag (Thailand).

Jun Xia and Siritree Suttajit: none known.

Sources of support

Internal sources

  • Chiang Mai University, Chiang Mai, Thailand.

External sources

  • No sources of support supplied

Differences between protocol and review

We have updated the methods section to reflect changes in the standard schizophrenia group template since publication of our protocol, for example inclusion of a 'Summary of findings' table. The methods published in the protocol are in Appendix 1.

We planned that two review authors would independently extract the data, however; due to language barrier, the Chinese data were extracted by one review author, JX.

One of the secondary outcomes (no clinical important change in global state) replicated the global state, which was the primary outcome, therefore, we deleted the duplicated outcome in the secondary outcomes and changed the title of the remaining outcome to be relapse.

We planned to enter skewed data from studies of less than 200 participants in additional tables rather than into the analyses. However, we found that most of the data from included studies were skewed (e.g. the data from all the studies investigated PANSS positive symptoms) and excluding all studies on the basis of estimates of the normal distribution would lead to selection bias. We therefore included all studies in the primary analysis and excluded the skewed data in the sensitivity analyses.

We planned to exclude studies with an attrition rate over 40%. However, many studies including vs HLP - Arvanitis 1997, vs FLUPHEN - Conley 2005, vs HLP - Fl'hacker 2005, vs PERPHEN - L'rman 2005, vs HLP - Purdon 2001 and vs HLP - Velligan 2002) reported high attrition rates over 40%. Moreover, it is still unclear what degree of attrition leads to a high degree of bias (Komossa 2010). We, therefore, did not exclude these studies on the basis of the attrition rate but we undertook sensitivity analyses of the main mental state outcomes, excluding studies with high attrition rates. We also addressed the attrition problems as well as the use of intention-to-treat in the 'Risk of bias' table, the results and the discussion.

We planned to carry out sensitivity analyses in different clinical groups. However, we did not find any difference in clinical groups, therefore, these sensitivity analyses were not done.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

vs CPZ - Ai 2007

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 12 weeks.
Setting: inpatients, Liaoning Renming Hospital, Liaoning Province, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 85.
Sex: male and female.
Age: mean 30 years, SD 11 years.
History: duration ill mean 5.8 years, SD 3.3 years.
Exclusion: severe physical illness.
Interventions1. Quetiapine: 400-750 mg/day. N = 43.
2. Chlorpromazine: 400-550 mg/day. N = 42.
Outcomes

Mental state: BPRS.

Global state: no clinical improvement*.

Adverse events.

Notes

*BPRS decreased rate < 25%.

BPRS = Brief Psychiatric Rating Scale.

CCMD = Chinese Classification of Mental Disorders.
TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS measured, but not reported.
Other biasLow riskNone obvious.

vs CPZ - AstraZeneca 2005

MethodsAllocation: randomised, no further details.
Blinding: double, no further details.
Duration: 10 weeks, preceded by 4-week single-blind haloperidol treatment phase and a 1-week washout period.
Design: parallel, multicentre.
Setting: USA and Canada.
ParticipantsDiagnosis: schizophrenia (DSM-IV), resistant to treatment with standard antipsychotic agents.
N = 260*.
Sex: male 206, female 54.
Age: 18-65 years (mean 40.9 years).
History: receiving haloperidol 40 mg/day for ≧ 2 weeks, < 30% reduction in BPRS, either CGI-S ≧ 4 or BPRS ≧ 10.
Interventions

1. Quetiapine: 600 mg/d. N = 130.

2. Chlorpromazine: 1200 mg/d. N = 130.

OutcomesLeaving the study early.
Global state: CGI-S.
Mental state: BPRS, Scale for the Assessment of Negative Symptoms (SANS).
Adverse effects: death, extrapyramidal side effects (AIMS, SAS).
Notes

*7 patients were excluded from the ITT analysis due to protocol deviation (5 in quetiapine group, 2 in chlorpromazine group.)

BPRS = Brief Psychiatric Rating Scale.

CGI-S = Clinical Global Impression - Severity scale.

AIMS = Abnormal Involuntary Movement Scale.

SAS = Simpson Angus scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble blind, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available but total was high (31.5%).
Last-observation-carried-forward method used (assumes that participants who discontinue remain stable).
Selective reporting (reporting bias)Low riskEvidence for selective reporting was not found.
Other biasHigh riskSponsored by manufacturers of quetiapine.

vs CPZ - Bai 2005

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 8 weeks.
Setting: Shanghai mental health centre, China.
ParticipantsDiagnosis: schizophrenia (ICD-10).
N = 105.
Age: mean 33 years, SD 11years.
Sex: not stated.
History: duration ill mean 49 months, SD 66 months.
Exclusion: severe physical illness, drug/alcohol dependent, received depot antipsychotic within 2 weeks prior to the trial.
Interventions1. Quetiapine: dosage not stated. N = 45.
2. Chlorpromazine: dosage not stated. N = 46.
OutcomesCognitive functioning: WAIS, WCST, WMS.
Unable to use: leaving the study early - no n number reported for individual groups.
Notes

ICD-10 = International Classification of Diseases Version 10.
PANSS = Positive and Negative Syndrome Scale.
WAIS-R = Wechsler Adult Intelligence Scale-Revised.

WCST = Wisconsin Card Sorting Test.

WMS = Wechsler Memory Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskno details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskno details.
Incomplete outcome data (attrition bias)
All outcomes
High riskITT was not used.
Selective reporting (reporting bias)High riskPANSS was not reported.
Other biasLow riskNone obvious.

vs CPZ - Cao 2005

MethodsAllocation: randomised, no further details.
Blinding: open label.
Duration: 16 weeks.
Setting: Inpatients, Nanjing brain hospital, Nanjing city, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 100.
Age: 25-40 years.
Sex: male.
History: duration ill, age of onset: not reported.
Exclusion: having other mental health problems, other physical illness, drug/alcohol. dependent, abnormal lab test results, problematic marital relationships.
Interventions1. Quetiapine: dosage not stated. N = 25.
2. Chlorpromazine: dosage not stated. N = 25.
Outcomes

Leaving the study early

Unable to use.
Adverse events - continuous data was reported, but the scale used was not clear.
Sexual arousal, and other sexual related outcomes that were measured using an invalidated scale.

Notes

Only male participants were included in the study; therefore, gender bias can
not be excluded.

CCMD = Chinese Classification of Mental Disorders.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
High riskOpen-label.
Incomplete outcome data (attrition bias)
All outcomes
High riskDrop outs were excluded from analysis.
Selective reporting (reporting bias)Low riskAll measured outcomes reported.
Other biasLow riskNone obvious.

vs CPZ - Chen 2001

MethodsAllocation: randomised using random number table.
Blinding: double, no further details.
Duration: 1 week wash out period + 8 weeks intervention.
Setting: inpatients, multi-centre, China.
ParticipantsDiagnosis: schizophrenia (CCMD-2-R).
N = 221.
Sex: male and female.
Age: 18-63 years.
History: duration ill, age of onset: not reported.
Exclusion: severe heart, renal, liver illness, nerve system illness, hypertension, blood illness, pregnant/lactating women, received ECT within 2 weeks prior to current study, suicidal patients, participated in other clinical trials within 1 month prior to current study.
Interventions1. Quetiapine: 200-800 mg/day. N = 114.
2. Chlorpromazine: 200-800 mg/day. N = 107.
Outcomes

Mental state: BPRS, PANSS.

Global state: no clinical improvement*.

Leaving the study early.
Adverse events.

Notes

*PANSS decreased rate < 25%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

BPRS = Brief Psychiatric Rating Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomised using random number table.
Allocation concealment (selection bias)Low riskPharmacist produced identical pills, only they know which pill contains experimental drug.
Blinding (performance bias and detection bias)
All outcomes
Unclear risk

Double, no further details.

Whether blinding was successful was not examined.

Incomplete outcome data (attrition bias)
All outcomes
Low riskITT was used.
Selective reporting (reporting bias)Low riskOutcomes reported as measured.
Other biasLow riskNone obvious.

vs CPZ - Chen 2007b

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 3 months.
Setting: inpatients and outpatients, Xuzhou, Jiangsu province, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 62.
Sex: male and female.
Age: 18-58 years.
History: duration ill: 6 months to 7 years.
Exclusion: severe physical illness, drug/alcohol dependent, pregnant/lactating women.
Interventions1. Quetiapine: 300.7 ± 100.3 mg/day. N = 32.
2. Chlorpromazine: 360.3 ± 82.5 mg/day. N = 30.
Outcomes

Mental state: PANSS, BPRS.

Adverse events.

Unable to use: GQOLI-74, only subscale scores available.

Notes

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

BPRS = Brief Psychiatric Rating Scale.

GQOLI = General Quality Of Life Inventory.
TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomisation, no further details.
Allocation concealment (selection bias)Unclear riskNot reported.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNot reported.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS was measured, but not reported.
Other biasLow riskNone obvious.

vs CPZ - Guo 2003

MethodsAllocation: randomised (by tossing a coin).
Blinding: not reported.
Duration: 8 weeks.
Setting: outpatients, Henan city, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
Sex: male and female.
Age: 18-55 years.
History: duration ill <1 year.
N = 148.
Exclusion: severe physical illness, drug/alcohol dependent.
Interventions1. Quetiapine: 450 ± 25 mg/day. N = 75.
2. Chlorpromazine: 600 ± 50 mg/day. N = 73.
Outcomes

Mental state: PANSS (total, positive and negative score).

Global state: no clinical improvement*

Adverse events.

Notes

*PANSS decreased rate <25%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomised by tossing a coin.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS scale score measured but not reported.
Other biasLow riskNone obvious.

vs CPZ - Guo 2007

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 8 weeks.
Setting: inpatients and outpatients, Kangning hospital, Guangzhou city, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 52.
Age: 18-40 years.
Sex: male and female.
History: duration ill 0.5-2 years.
Exclusion: severe physical illness, drug/alcohol dependent, pregnant/lactating women.
Interventions1. Quetiapine: 600-800 mg/day. N = 26.
2. Chlorpromazine: 300-500 mg/day. N = 26.
Outcomes

Mental state: PANSS.

Global state: CGI-S, CGI-I.

Adverse events

Notes

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

CGI-S = Clinical Global Impression - Severity scale.

CGI-I = Clinical Global Impression - Improvement scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS was measured, but not reported.
Other biasLow riskNone obvious.

vs CPZ - He 2003

MethodsAllocation: randomised, no further details,
Blinding: open label.
Duration: 6 weeks.
Setting: Jiangsu City, China.
ParticipantsDiagnosis: schizophrenia (CCMD-2-R).
Sex: male and female.
Age: 18-50 years.
N = 40.
History: duration ill: 1 month - 10 years.
Exclusion criteria: severe physical illness, alcohol/drug abuse, pregnant/lactating women.
Interventions1. Quetiapine: 433 ± 19 mg/day. N = 20.
2. Chlorpromazine: 468 ± 39 mg/day. N = 20.
OutcomesMental state: PANSS endpoint score.
Global state: no clinical improvement*.
Adverse events.
Notes

*PANSS decreased rate <40%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
High riskOpen label.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS score measured, but not reported.
Other biasLow riskNone obvious.

vs CPZ - Hu 2003

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 4 weeks.
Setting: inpatients, Mental health centre, Sichuan, China.
ParticipantsDiagnosis: schizophrenia (DSM-IV), first onset, inpatients.
N = 62.
Sex: male
Age: mean 26.94, SD 8.82
History: duration ill, age of onset: not reported.
Exclusion: severe physical illness, lactating/pregnant women.
Interventions1. Quetiapine: 410 ± 108.77 mg/d. N = 22.
2. Chlorpromazine: 350 ± 67.8 mg/d. N = 19.
Outcomes

Mental state: PANSS (positive, negative, general psychopathology subscale.)

Adverse events: TESS, RSESE.

Unable to use: GAF score - some data missing.

Notes

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

RSESE = Rating Scale for Extrapyramidal Side Effects.

GAF = Global Assessment of Functioning.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was no found.
Other biasLow riskNone obvious.

vs CPZ - Ji 2004

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 3 months.
Setting: inpatients, FuZhou mental illness prevention hospital, FuJian province, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 66.
Age: 18-45 years.
Sex: male and female.
History: duration ill mean 1.6 years, SD 0.5 years.
Exclusion: patients with severe physical illnesses, drug or alcohol dependency, pregnant or lactating women.
Interventions1. Quetiapine: dosage not stated. N = 33.
2. Chlorpromazine: dosage not stated. N = 33.
Outcomes

Leaving the study early.

Mental state.
PANSS total, negative, positive, general psychopathology score
Quality of life GQOLI-74 score.

Adverse effect.

Notes

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

GQOLI = General Quality Of Life Inventory.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
High riskDrop outs are excluded from analysis.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasLow riskNone obvious.

vs CPZ - Jiang 2006

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 8 weeks.
Setting: inpatients, Shanghai mental health centre, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 191.
Age: mean 34 years, SD 9.7 years.
Sex: not reported.
History: duration ill mean 58 months, SD 68 months.
Exclusion: severe physical illness, drug/alcohol dependent, received experimental drugs within 4 weeks prior to the trial.
Interventions1. Quetiapine: 100-800 mg/d. N = 94.
2. Chlorpromazine: 100-600 mg/d. N = 97.
Outcomes

Leaving the study early.

Adverse events.

Notes

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskDrop outs were included in final analysis.
Selective reporting (reporting bias)High riskPANSS measured, but not reported.
Other biasLow riskNone obvious.

vs CPZ - Jin 2007

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 4 weeks.
Setting: outpatients and inpatients, Hebei province, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 60.
Sex: male and female.
Age: 18-60 years.
History: duration ill 1-24 months.
Exclusion: severe physical illness, drug abuse, pregnant/lactating women, allergic to either of the intervention drugs.
Interventions1. Quetiapine: 200-800 mg/day. N = 30.
2. Chlorpromazine: 200-600 mg/day. N = 30.
Outcomes

Mental state: PANSS score.

Global state: no clinical improvement*

Adverse events.

Notes

*PANSS decreased rate <25%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS measured, but not reported.
Other biasLow riskNone obvious.

vs CPZ - Li 2003

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 6 months.
Setting: inpatients and outpatients, mental health centre, Shandong, China.
ParticipantsDiagnosis: schizophrenia (CCMD-2-R).
N = 103.
Sex: male and female.
Age: 18-56 years.
History: duration ill mean 1.2, SD 2.1 years.
Exclusion: severe physical illness, drug/alcohol dependent, pregnant/lactating women.
Interventions1. Quetiapine: 482 ± 124 mg/day. N = 52.
2. Chlorpromazine: 503 ± 134 mg/day. N = 51.
Outcomes

Leaving the study early.

Mental state: PANSS scores.

Adverse effects: TESS scores.

Unable to use: WHO QOL-100 subscale scores.

Notes

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

WHO QOL = World Health Organization Quality of Life.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
High riskITT was not used.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasLow riskNone obvious.

vs CPZ - Link 1997

Methods

Allocation: randomised, no further details.
Blinding: investigator-blinded, no further details.
Duration: 6 weeks, preceded by 24 hours washout period.
Design: parallel, multicentre.
Setting: Belgium, UK, Spain, France, South Africa.

Consent: obtained.

Loss: described.

Participants

Diagnosis: schizophrenia, schizophreniform disorder (DSM-III-R).
N = 201.
Sex: male 129, female 72.
Age: 18 to 65 years.
History: BPRS ≧ 27, at least 3 on two or more of the BPRS positive symptom items 'conceptual disorganisation', 'suspiciousness', hallucinatory behaviour' and 'unusual thought content'; CGI-S ≧4.

Exclusion: medical condition or laboratory abnormality that might confound the trial results or put the patient at risk, receiving long-acting depot medication, participated in another investigational drug trial during the 4 weeks prior to randomisation, significant alcohol or other drug abuse within the previous 12 months.

Interventions

1. Quetiapine: mean 407 mg/day; maximum 750 mg/day (but not more than 500 mg/day for more than 14 days). N = 101.

2. Chropromazine: mean 384 mg/day; maximum 750 mg/day. N = 100.

OutcomesLeaving the study early.
Global state: CGI (no SD).
Mental state: PANSS (no raw data), BPRS (no SD).
Adverse effects: extrapyramidal side effects (AIMS, BAS, SAS), haematology, liver function test, thyroid function tests, ECGs.
Notes

PANSS = Positive and Negative Syndrome Scale.

BPRS = Brief Psychiatric Rating Scale.

CGI = Clinical Global Impression.

AIMS = Abnormal Involuntary Movement Scale.

BAS = Barnes Akathisia Scale.

SAS = Simpson Angus scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble-blinded, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available.
ITT method was not used but overall discontinuation rate was low (2.5%).
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasHigh riskSponsored by manufacturer of quetiapine.

vs CPZ - Mei 2007

MethodsAllocation: randomised, no further details..
Blinding: not reported.
Duration: 8 weeks.
Setting: inpatients, Rongjun hospital, Anhui, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 96.
Sex: male and female.
Age: 17-60 years.
History: duration ill 3 months to 12 years.
Exclusion: severe physical illness, drug/alcohol dependent, pregnant/lactating women.
Interventions1. Quetiapine: 600 ± 50 mg/day. N = 47.
2. Chlorpromazine: 425 ± 106 mg/day. N = 49.
Outcomes

Mental state: PANSS.

Global state: no clinical improvement*.

Adverse events.

Notes

*PANSS decreased rate <25%

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.
TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS measured, but not reported.
Other biasLow riskNone obvious.

vs CPZ - Peng 2006

MethodsAllocation: randomised using random number table.
Blinding: double, no further details.
Duration: 42 days.
Setting: inpatients, West China Hospital, Sichuan, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 40.
Age: mean 30.6 years, SD 11.26 years.
Sex: male and female
History: duration ill mean 28.5 years, SD 6.45years
Exclusion: not stated.
Interventions1. Quetiapine: 555.56 ± 137.74 mg/day, N = 20.
2. Chlorpromazine: 473.61 ± 107.26 mg/day, N = 20
Outcomes

Leaving the study early.

Unusable data:
PANSS, BPRS, TESS, RSESE scores - no SD reported.

Notes

PANSS = Positive and Negative Syndrome Scale.

BPRS = Brief Psychiatric Rating Scale.

TESS = Treatment Emergent Symptom Scale.

RSESE = Rating Scale for Extrapyramidal Side Effects.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomised using random number table.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble blind, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskUnclear whether people dropped out are included in the analysis.
Selective reporting (reporting bias)Low riskNo evidence of selective reporting.
Other biasLow riskNone obvious.

vs CPZ - Tian 2006

MethodsAllocation: randomised, no further details..
Blinding: not reported.
Duration: 8 weeks.
Setting: inpatients and outpatients, Xiangfan No 1 people's hospital, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 92.
Sex: male and female.
Age: 18-61 years.
History: duration ill mean 6 years, SD 4 years.
Exclusion: severe physical illness, drug/alcohol dependent, pregnant/lactating women, aggressive and suicidal patients.
Interventions1. Quetiapine: 421±18 mg/day. N = 46.
2. Chlorpromazine: 420±67.2 mg/day. N = 46.
Outcomes

Mental state: PANSS scores.

Global state: no clinical improvement.

Adverse events.

Notes

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasLow riskNone obvious.

vs CPZ - Wang 2004

MethodsAllocation: randomised, no further details.
Blinding: no details.
Duration: 6 weeks.
Setting: inpatients, Guangzhou Mental Health Hospital, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 96.
Sex: male and female.
Age: 19-58 years.
History: duration ill 2 months - 14 years.
Exclusion criteria: pregnant/lactating women, drug/alcohol dependent patients, severe physical illness, took quetiapine or chlorpromazine within 4 weeks prior to the study, suicidal, allergic to any of the intervention drugs, participated in other clinical trials within 30days prior to the current study.
Interventions1. Quetiapine: 300-750 mg/day. N = 48.
2. Chlorpromazine: 300-750 mg/day. N = 48.
Outcomes

Leaving the study early.

Mental state: PANSS (total, positive, negative,general psychopathology subscale).

Global state: CGI endpoint total improvement score, illness severity score, no clinical improvement*.

Adverse events

Notes

*PANSS decreased rate < 30%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

CGI = Clinical Global Impression.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskITT was used.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasLow riskNone obvious.

vs CPZ - Wang 2005

MethodsAllocation: randomised,no further details..
Blinding: open label.
Duration: 8 weeks.
Setting: inpatients, Shanghai mental health centre, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 105.
Age: mean 33 years, SD 11 years.
Sex: not stated.
History:duration ill mean 65 days, SD 8 days.
Exclusion: drug/alcohol dependent patients, severe physical illness, took quetiapine or chlorpromazine within 4 weeks prior to the study.
Interventions1. Quetiapine: 400-750 mg/day. N = 48.
2. Chlorpromazine: 100-600 mg/day. N = 57.
Outcomes

Leaving study early.

Mental state: PANSS scores.

Global state: no clinical improvement*.

Adverse effect.

Notes

*PANSS decreased rate <25%

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomisation, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
High riskopen label.
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo details.
Selective reporting (reporting bias)High riskTESS score measured not reported.
Other biasLow riskNone obvious.

vs CPZ - Zhang 2002

MethodsAllocation: randomised (by tossing a coin).
Blinding: double blind, no further details.
Duration: 8 weeks.
Setting: inpatients, RiZhao mental health centre, Shandong Province, China.
Participants

Diagnosis: schizophrenia (CCMD-2-R).

N = 117.
Sex: male and female.
Age: mean age 24 years, SD 5 years.
History: duration ill mean 3.8 years, SD 1.1 years.
Exclusion criteria: length of illness > 5 years, with severe physical or neurological illness, pregnant or lactating women.

Interventions1. Quetiapine: 300 mg-550 mg/day. N = 60.
2. Chlorpromazine: 300 mg-550 mg/day. N = 57.
Outcomes

Mental state: PANSS positive, negative, general psychopathology and endpoint score and change score.
Global state: no clinical improvement*.
Adverse event: TESS endpoint and change score.

Unusable data:
Leaving the study early, no 'n' number reported.

Notes

* PANSS decreased rate < 25%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomised by tossing a coin.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear risk

Double blind, no further details.

Whether blinding was successful was not examined.

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasLow riskNone obvious.

vs CPZ - Zhang 2003

MethodsAllocation: randomised using computer generated random number.
Blinding: double, no further details.
Duration: 1 week wash out period + 6 weeks of intervention + 1 week with reduced dose of medication to help patients to withdraw from the interventions.
Setting: inpatients, multicentre, China.
ParticipantsDiagnosis: schizophrenia (CCMD-2-R).
N = 237.
Sex: male and female.
Age: mean age 32 years, SD 11.
History: duration ill: mean Quetiapine 7.0 ± 7.3 years, mean Chlorpromazine 5.9±7.2 years.
Exclusion criteria: patients with neurological, heart, liver, renal or other severe physical illnesses; pregnant or lactating women; alcohol or drug dependent patients; patients who received quetiapine or chlorpromazine 4 weeks prior to the study; patients with allergy of quetiapine and chlorpromazine; patients who participated in other clinical drug trials within the last 30 days of this study.
Interventions1.Quetiapine: first week 75mg-300 mg/day, second week onwards 300mg-750 mg/day. N = 119.
2.Chlorpromazine:first week 75mg-300 mg/day, second week onwards 300mg-750 mg/day. N = 118.
Outcomes

Leaving the study early.
Mental state, PANSS end point score and PANSS score decreased rate.

Unusable data:
CGI score measured, but data not reported.
No clinical improvement: no n number reported.
Adverse effect measured, but 'n' numbers were not reported.

Notes

CCMD = Chinese Classification of Mental Disorders.

CGI =

PANSS = Positive and Negative Syndrome Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomised using computer-generated random number.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear risk

Double blind, no further details.

Whether blinding was successful was not examined.

Incomplete outcome data (attrition bias)
All outcomes
Low riskITT was used.
Selective reporting (reporting bias)High riskCGI and adverse events were measured, but data not reported.
Other biasLow riskNone obvious.

vs CPZ - Zhong 2005

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 12 weeks.
Setting: inpatients, Beijing hospital, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 84.
Sex: not stated.
Age: 16-60 years.
Exclusion: patients with severe physical illness, or drug/alcohol dependent.
Interventions1.Quetiapine: 400-900 mg/day. N = 42.
2.Chlorpromazine: 300-800 mg/day. N = 42.
Outcomes

Mental state: PANSS negative, positive, general psychopathology, total score.
Global state: SF-36 total score.

Adverse events.

Notes

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

SF-36 = Short Form (36) Health Survey.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS measured, but not reported.
Other biasLow riskNone obvious.

vs CPZ - Zhou 2003

MethodsAllocation: randomised, using random number table.
Blinding: assessors are blinded, unclear if the patients are.
Duration: 6 weeks.
Setting: inpatients, Mental health centre, Sichuan, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3), inpatients.
N = 40.
Sex: male and female.
Age: mean˜28, SD˜10 years.
History: duration ill mean˜25, SD˜6 years.
Exclusion: not described.
Interventions1. Quetiapine: dosage not stated. N = 20.
2. Chlorpromazine: dosage not stated. N = 20.
Outcomes

Leaving study early.

Unable to use: WCST subscale scores,

Notes

CCMD = Chinese Classification of Mental Disorders.

WCST = Wisconsin Card Sorting Test.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomised using random number table.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskBlinding, but unclear whether single or double blind (assessors are blinded anyway).
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasLow riskNone obvious.

vs CPZ - Zhou 2004

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 12 weeks.
Setting: inpatients, Jinan psychiatric hospital, Shandong province, China.
ParticipantsDiagnosis: schizophrenia (ICD-10, CCMD-3).
N = 83.
Age: 17-51 years old.
Sex: male and female.
History: duration ill 4 months to 21 years.
Exclusion criteria: not stated.
Interventions1. Quetiapine: 300-400 mg/d. N = 42.
2. Chlorpromazine: 200-600 mg/d. N = 41.
Outcomes

Leaving the study early.

Mental state: PANSS positive, negative, general psychopathology, total score.

Global state: no clinical improvement*

Quality of life: GQOLI score.

Adverse events: TESS score.

Notes

*PANSS decreased rate <25%

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

GQOLI = General Quality Of Life Inventory.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNo n number reported, so not clear what happened to people who dropped out.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not clear.
Other biasLow riskNone obvious.

vs CPZ - Zou 2006

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 8 weeks.
Setting: Community and hospital, Zhejiang, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 86.
Age: 17-56 year.
Sex: male and female.
History: duration ill <10 years.
Exclusion: severe physical illness, drug/alcohol dependent, pregnant/lactating women, allergic to experimental drugs.
Interventions1. Quetiapine: 483.1 ± 25.5 mg/day. N = 43.
2. Chlorpromazine: 450.2 ± 30.2 mg/day. N = 43.
Outcomes

Mental state: PANSS scores.

Global state: no clinical improvement*.

Adverse events.

Notes

*PANSS decreased rate <25%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS scores measured, but not reported.
Other biasLow riskNone obvious.

vs FLUPHEN - Conley 2005

Methods

Allocation: randomised, no further details.
Blinding: double, no further details.
Duration: 12 weeks, preceded by 4-6 weeks of open-label trial with olanzapine or typical antipsychotic other than fluphenazine.
Design: parallel.
Setting: inpatients, USA.

Consent: obtained.

Loss: described.

ParticipantsDiagnosis: schizophrenia (DSM-IV).
N = 52.*
Sex: male 44, female 8.
Age: 18-65 years (mean 44 years).
History: BPRS ≧ 4 on 2 of 4 psychosis items, BPRS total ≧ 45, CGI ≧ 4, 2 prior failed treatment with 2 antipsychotics, no stable period of good social/occupational functioning within 5 years.
Interventions

1. Quetiapine: 300-500 mg/d. N = 12.

2. Fluphenazine: 10-15 mg/d. N = 13.

Outcomes

Global state: CGI-S, quality of life.

Mental state: BPRS.

Adverse effects: Extrapyramidal side effects (SAS, AIMS), Barnes Akathisia Scale.

Notes

BPRS = Brief Psychiatric Rating Scale.

CGI-S = Clinical Global Impression - Severity scale.

AIMS = Abnormal Involuntary Movement Scale.

SAS = Simpson Angus scale.

*Thirteen participants were randomly assigned to risperidone.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
High riskData for leaving study early were available but total was high (47%).
Last-observation-carried-forward method was not used.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasHigh riskSponsored by manufacturers of quetiapine and risperidone.

vs HLP - Arvanitis 1997

Methods

Allocation: randomised, no further details.
Blinding: double, no further details.
Duration: 6 weeks, preceded by 7 days washout period.*
Design: parallel, multicentre.
Setting: inpatients, USA and Canada.

Consent: obtained.

Loss: described.

Participants

Diagnosis: schizophrenia (DSM-III-R).
N = 361.
Sex: male 274, female 87.
Age: 18-65 years (mean = 37 years).
History: BPRS ≧ 27, 3 on two or more of the BPRS positive symptom items 'conceptual disorganisation', 'suspiciousness', hallucinatory behaviour' and 'unusual thought content'; CGI-S ≧ 4.

Exclusion: concurrent Axis I DSM-III-R diagnoses; history of seizure disorder; clinically significant medical condition; participation in another investigational drug trial within 30 days of trial entry; use of depot antipsychotics within one dosing interval; pregnancy.

Interventions

1. Quetiapine: 75 mg/day. N = 53.

2. Quetiapine: 150 mg/day. N = 48.

3. Quetiapine: 300 mg/day. N = 52.

4. Quetiapine: 600 mg/day. N = 51.

5. Quetiapine: 750 mg/day. N = 54.

2. Haloperidol:12 mg/day. N = 52.

OutcomesLeaving the study early.
Global state: CGI.
Mental state: BPRS, Modified Scale for the Assessment of Negative Symptoms (SANS).
Adverse effects: extrapyramidal side effects (AIMS, SAS), haematology, liver function test, serum prolactin (no SD), thyroid function tests, weight, ECGs.
Notes

*After the 7-day, single blind, placebo washout phase, patients with a 20% or greater decrease in BPRS total score or a greater than 1-point decrease in CGI-S was considered placebo responders and were withdrawn.

BPRS = Brief Psychiatric Rating Scale.

AIMS = Abnormal Involuntary Movement Scale.

SAS = Simpson Angus scale.

ECG = electrocardiogram.

CGI-S = Clinical Global Impression - Severity scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available but total was high (59%).
Last-observation-carried-forward method used (assumes that participants who discontinue remain stable).
Selective reporting (reporting bias)Low riskEvidence for selective reporting was not found.
Other biasHigh riskSponsored by manufacturers of quetiapine.

vs HLP - Atmaca 2002

Methods

Allocation: randomised, no further details.
Blinding: open label.
Duration: 6 weeks, preceded by 14 days washout period.
Design: single centre.
Setting: inpatient and outpatient, Turkey.

Consent: obtained.

Loss: described (all participants completed the study.)

Participants

Diagnosis: schizophrenia (DSM-IV).
N = 35.
Sex: female 35.
Age: 18-45 years (mean quetiapine = 27.62 years and mean haloperidol = 29.44 years).
History: duration ill and age of onset: not reported.

Exclusion: severe physical illness, history of alcohol and substance abuse or dependence, presence of any endocrinology state, taking oral contraceptives.

Interventions

1. Quetiapine: 600 mg/day. N = 18.

2. Haloperidol: 10 mg/day. N = 17.

OutcomesLeaving the study early: any cause, lack of efficacy, intolerability, patient's decision.
Mental state: PANSS, BPRS.
Adverse effects: extrapyramidal side effects (ESRS), prolactin level.
Notes

PANSS = Positive and Negative Syndrome Scale.

BPRS = Brief Psychiatric Rating Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
High riskOpen label; however, some outcome measures such as prolactin level are unlikely to be influenced by lack of blinding.
Incomplete outcome data (attrition bias)
All outcomes
Low riskAll participants completed the study.
Selective reporting (reporting bias)Low riskEvidence for selective reporting was not found.
Other biasHigh riskOnly female participants could be included in the study; therefore, gender bias can not be excluded.

vs HLP - Copolov 2000

Methods

Allocation: randomised, no further details.
Blinding: double, no further details.
Duration: 6 weeks, preceded by 48 hours washout period.
Design: parallel, multicentre.
Setting: inpatient, 14 countries.

Consent: obtained.

Loss: described.

Participants

Diagnosis: Schizophrenia (DSM-III-R).
N = 448.
Sex: male 305, female 143.
Age: 18 years or over (mean quetiapine 37 years, mean haloperidol 37 years).
History: acute exacerbation of chronic or subchronic schizophrenia; CGI-S ≧ 4; PANSS ≧ 60, at least 4 on two or more of the PANSS items 'delusions', 'conceptual disorganization', 'hallucinatory behaviour', and 'suspiciousness/persecution'.

Exclusion: significant clinical disorders or laboratory or ECG findings, epilepsy, low WBC count; pregnant or lactating women, or women of reproductive age not using adequate contraception; treatment with a long-acting depot medication less than one dosing period; participation in any other investigational drug trial during 4 weeks prior to randomisation.

Interventions

1. Quetiapine: range 300-600 mg/day, maximum 800 mg/day. N = 221.

2. Haloperidol: range 6-12 mg/day, maximum 16 mg/day. N = 227.

OutcomesLeaving the study early.
Global state: CGI.
Mental state: PANSS.
Adverse effects: extrapyramidal side effects (AIMS, SAS), haematology, liver function test, serum prolactin, thyroid function tests, weight.
Notes

PANSS = Positive and Negative Syndrome Scale.

CGI = Clinical Global Impression.

AIMS = Abnormal Involuntary Movement Scale.

SAS = Simpson Angus scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available but total was high (33%).
Last-observation-carried-forward method used (assumes that participants who discontinue remain stable).
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasHigh riskSponsored by manufacturers of quetiapine.

vs HLP - Emsley 1999

Methods

Allocation: randomised, no further details.
Blinding: double, no further details.
Duration: 8 weeks, preceded by 4 weeks fluphenazine run-in phase.
Design: parallel, multicentre.
Setting: Europe and South Africa.

Consent: obtained.

Loss: described.

Participants

Diagnosis: Schizophrenia (DSM-IV).
N = 365*.
Sex: male 203, female 85.
Age: 18 years or over.
History:persistent positive symptoms while taking antipsychotics; positive scale of PANSS ≧ 15, at least 4 on two or more of the PANSS items 'delusions', 'conceptual disorganization', 'hallucinatory behaviour', and 'suspiciousness/persecution'; CGI-S ≧ 3.

Exclusion: clozapine non-responders; acute exacerbation of schizophrenia within the previous 3 months; known sensitivity to drugs evaluated in the trial or history of idiopathic or drug-induced agranulocytosis.

Interventions

1. Quetiapine: 600 mg/day. N = 143.

2. Haloperidol: 20 mg/day. N = 145.

Outcomes

Leaving the study early (death, adverse event, non-compliance with protocol, deterioration of condition, informed consent withdrawn, patient lost to follow-up).

Unusable data:

Global state: CGI (no SD).
Mental state: PANSS (positive, negative, general psychopathology subscale) (no SD), BPRS (total, positive subscale, mood cluster) (no SD).
Adverse effects: extrapyramidal side effects (SAS), haematology, serum prolactin (no SD).

Notes

*After 4 weeks run-in phase, 288 participants were included and randomised.

PANSS = Positive and Negative Syndrome Scale.

BPRS = Brief Psychiatric Rating Scale.

CGI = Clinical Global Impression.

SAS = Simpson Angus scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble blind, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available.
Last-observation-carried-forward method used (assumes that participants who discontinue remain stable).
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasHigh riskSponsored by manufacturers of quetiapine.

vs HLP - Emsley 2004

Methods

Allocation: randomised, no further details.
Blinding: investigator-blinded, no further details.
Duration: 50 weeks, preceded by 2 weeks washout period.
Design: parallel, single centre.
Setting: inpatients and outpatients, South Africa.

Consent: obtained.

Loss: described.

Participants

Diagnosis: Schizophrenia, schizoaffective disorder (DSM-IV).
N = 47*.
Sex: male 29, female 16.
Age: 18 to 65 years (mean quetiapine 49.2 years, mean haloperidol 50.1 years)
History:diagnosis of tardive dyskinesia (Schooler and Kane criteria).

Exclusion: resistant to standard antipsychotic medication; clozapine non-responders; acute exacerbation of schizophrenia within the previous 3 months; known sensitivity to drugs evaluated in the trial or history of idiopathic or drug-induced agranulocytosis.

Interventions

1. Quetiapine: 100 mg/day for 2 days, 200 mg/day for 2 days, 300 mg/day for 2 days, 400 mg/day from day 7 onwards; maximum 800 mg/day. N = 22.

2. Haloperidol: 5 mg/day for 4 days, 10 mg/day for 3 days and onwards; maximum 20 mg/day. N = 23.

OutcomesMental state: PANSS (positive, negative, general psychopathology subscale).
Adverse effects: dyskinesia (ESRS dyskinesia subscale), weight, BMI, serum prolactin, glycosylated haemoglobin.
Notes

*Two were excluded (1 withdrew before reaching the target dose, 1 had unrelated medical illness).

PANSS = Positive and Negative Syndrome Scale.

ESRS = Extrapyramidal Symptom Rating Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskSingle-blinded (investigator-blinded), no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available but total was high (38%).
Last-observation-carried-forward method used (assumes that participants who discontinue remain stable).
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasHigh risk

ESRS dyskinesia subscale and CGI dyskinesia scores were not shown in the result.

Sponsored by manufacturer of quetiapine.

vs HLP - Fl'hacker 2005

Methods

Allocation: randomised, web-based online system.
Blinding: open label.
Duration: 12 months.
Design: parallel, multicentre.
Setting: 13 European countries and Israel.

Consent: obtained.

Loss: described.

Participants

Diagnosis: schizophrenia, schizoaffective disorder, schizophreniform disorder (DSM-IV).
N = 498.
Sex: male 298, female 200.
Age: 18 to 40 years (mean quetiapine 26.4 years, mean haloperidol 25.4 years).
History: duration of illness less than 2 years; exposure to antipsychotic treatment not more than 2 weeks in the year preceding study entry or 6 weeks in lifetime.

Exclusion: known to be intolerance to one of the study drugs; met any of the contraindications for any of the study drugs.

Interventions

1. Quetiapine: range 200-750 mg/day. N = 104.

2. Haloperidol: range 1-4 mg/day. N = 103.

3. Olanzapine: range 5-20 mg/day. N = 105.

4. Ziprasidone: range 40-160 mg/day. N = 82.

5. Amisulpride: range 200-800 mg/day. N = 104.

Outcomes

Leaving the study early.

Global state: CGI, GAF.

Mental state: PANSS (positive, negative, general psychopathology subscale), CDSS.

Quality of life: Manchester short assessment of quality of life scale (MANSA)

Cognition: Composite score from Rey Auditory Verbal Learning test, Trail Making Test, WAIS Digit Symbol Test and Purdue Pegboard Test.
Adverse effects: Extrapiramidal side effect (ST. Hans rating scale; SHRS), dyskinesia (ESRS dyskinesia subscale), sexual dysfunction (selected items from the Udvalg for Kliniske Undersøgelser; UKU), weight, BMI, serum prolactin, glycosylated haemoglobin, ECG.

Notes

CGI = Clinical Global Impression.

GAF = Global Assessment of Functioning.

PANSS = Positive and Negative Syndrome Scale.

CDSS = Calgary Depression Scale for Schizophrenia.

WAIS = Wechsler Adult Intelligence Scale.

ESRS = Extrapyramidal Symptom Rating Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomised, web-based online system.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
High riskOpen label.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available but total was high (42%).
Last-observation-carried-forward method used (assumes that participants who discontinue remain stable).
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasHigh riskSponsored by manufacturer of quetiapine and other drug companies.

vs HLP - Huang 2007

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 18 months.
Setting: inpatients and outpatients, Guangzhou mental health hospital, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 128.
Age: 18-60 years.
Sex: male and female.
History: duration ill mean 5.6 years, SD 3 years.
Exclusion: severe physical illness, children, pregnant/lactating women, drug/alcohol dependent, suicidal, aggressive and agitated, allergic to intervention drugs, taking antipsychotic medication within 1 week prior to the trial.
Interventions1. Quetiapine: 400-700 mg/day. N = 64.
2. Haloperidol: 4-40 mg/day. N = 64.
Outcomes

Mental state: PANSS score.

Global state: no clinical improvement*

Adverse events.

Notes

*PANSS decreased rate <25%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete outcome data.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasLow riskNone obvious.

vs HLP - McCue 2006

Methods

Allocation: randomised, using randomisation website.
Blinding: open trial.
Duration: 3 weeks.
Design: single centre.
Setting: inpatients, USA.

Consent: obtained.

Loss: described.

Participants

Diagnosis: Schizophrenia, schizoaffective disorder, schizophreniform disorder (DSM-IV).
N = 327*.
Sex: male 198, female 121.
Age: 18 years and older.
History: duration ill: mean quetiapine 14.5 years, mean haloperidol 12.2 years.

Exclusion: pregnant or lactating; medical condition in which pharmacotherapy would prove a risk; history of response or lack of response to a particular antipsychotic drug; psychiatrist judged that would best not entered into the study; diagnosis of bipolar disorder, major depressive disorder, substance-induced psychotic disorder.

Interventions

1. Quetiapine: range 50-1200 mg/day; mean 652.5 mg/day. N = 50.

2. Haloperidol: range 4-30 mg/day; mean 16.0 mg/day. N = 57.

3. Aripiprazole: range 10-45 mg/day; mean 21.8 mg/day. N = 53.

4. Olanzapine: range 5-40 mg/day; mean 19.1 mg/day. N = 52.

5. Risperisdone: range 2-9 mg/day; mean 5.2 mg/day. N = 57.

6. Ziprasidone: range 40-240 mg/day; mean 151.2 mg/day. N = 50.

Outcomes

Effectiveness defined as no longer required acute in-patient care.

Mental state: BPRS.

Notes

*Eight participants were withdrawn for reasons unrelated to antipsychotic treatment and were not included in the analysis.

BPRS = Brief Psychiatric Rating Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRamdomised, using computer random number generator.
Allocation concealment (selection bias)Low riskStaff and patients did not have assess to the list.
Blinding (performance bias and detection bias)
All outcomes
High riskOpen label.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available.
ITT method was not used but overall discontinuation rate was low (5.6%).
Selective reporting (reporting bias)Low riskNo evidence of selective reporting was found.
Other biasLow riskNone obvious.

vs HLP - Murasaki 1999

Methods

Allocation: randomised, no further details.
Blinding: double, no further details.
Duration: 8 weeks.
Design: parallel, multicentre.
Setting: inpatient and outpatient, Japan.

Loss: described.

Participants

Diagnosis: Schizophrenia (ICD-10).
N = 197.
Sex: male 129, female 68.
Age: 18-65 years.
History: duration ill: mean quetiapine 17.5 years, mean haloperidol 16.9 years.

Exclusion: significant deterioration of personality, receiving high-dose antipsychotics, severe endocrinological disease, epilepsy or convulsive disorders, Parkinson's disease, organic cerebral disorder.

Interventions

1. Quetiapine: range 150-600 mg/day, mean 226 mg/day. N = 100.

2. Haloperidol: range 4.5-18 mg/day, mean 6.7 mg/day. N = 97.

OutcomesLeaving the study early.
Global state: Global Improvement Rating (GIR).
Mental state: PANSS (total, positive, negative and general psychopathology), BPRS.
Adverse effects: reported EPS, Overall Safety Rating (OSR), Global Usefulness Rating (GUR).
Notes

PANSS = Positive and Negative Syndrome Scale.

BPRS = Brief Psychiatric Rating Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
High riskData for leaving study early were available but total was high (39%).
ITT method was not used.
Selective reporting (reporting bias)Low riskNo evidence of selective reporting was found.
Other biasHigh riskSponsored by manufacturer of quetiapine.

vs HLP - Purdon 2001

Methods

Allocation: randomised, no further details.
Blinding: Single-blinded (investigator-blinded), no further details.
Duration: 6 months, preceded by 2 days washout period.
Design: multicentre.

Consent: obtained.

Loss: described.

Participants

Diagnosis: schizophrenia (DSM-IV).
N = 25.
Sex: male 20, female 5.
Age: mean 33.9 (SD 7.3) years.
History: duration ill and age of onset: not reported.

Exclusion: history of a serious medical disease or neurological disorder (including serious head injury); active substance abuse in the 30-day period before enrolment.

Interventions

1. Quetiapine: range 300-600 mg/day. N = 13.

2. Haloperidol: range 10-20 mg/day. N = 12.

Outcomes

Mental state: PANSS (positive, negative, general psychopathology subscale); CDS; BDI.

Cognition: Battery of tests used to determine functioning in 6 domains (motor skill, attention span, verbal fluency and reasoning, visuospatial fluency and construction) consists of finger tapping test, grooved pegboard test, Wechsler memory scale-revised, controlled oral word association test, WAIS-R similarities, design fluency, visual organisation test, complex figure copy test, trailmaking part B WAIS-R digit symbol test, Wisconsin card sorting test, Ray and Crawfoord auditory verbal learning tests, serial design learning tests, story recall test, Wechsler memory scale visual reproduction test, Reay and Taylor complex figure immediate recall test.

Adverse effects: EPS (AIMS, SAS).

Notes

PANSS = Positive and Negative Syndrome Scale.

CDS = Calgary Depression Scale.

BDI = Beck Depression Inventory.

WAIS-R = Wechsler Adult Intelligence Scale -Revised.

AIMS = Abnormal Involuntary Movement Scale.

SAS = Simpson Angus scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble-blinded, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available but total was high (48%).
Last-observation-carried-forward method used (assumes that participants who discontinue remain stable).
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasHigh riskSponsored by manufacturer of quetiapine.

vs HLP - Taneli 2003

Methods

Allocation: randomised, no further details.
Blinding: open label.
Duration: 12 weeks.
Design: parallel, multicentre.
Setting: outpatient, Turkey.

Consent: obtained.

Loss: described.

Participants

Diagnosis: schizophrenia (DSM-IV).
N = 79.
Sex: male 68, female 11.
Age: 18-65 years.
History: exacerbation of schizophrenia with PANSS ≧ 60; CGI-S ≧ 3.

Exclusion: no detail.

Interventions

1. Quetiapine: range 400-750 mg/day. N = 45.

2. Haloperidol: range 5-20 mg/day. N = 34.

Outcomes

Leaving the study early.

Global state: CGI-S.

Mental state: PANSS (positive, negative, general psychopathology subscale).

Adverse effects: extrapyramidal side effects (AIMS, SAS,ASC-SR).

Notes

PANSS = Positive and Negative Syndrome Scale.

CGI = Clinical Global Impression.

AIMS = Abnormal Involuntary Movement Scale.

SAS = Simpson Angus scale.

ASC-SR = Approaches to Schizophrenia Communication Self-Report.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
High riskOpen label.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available.
Last-observation-carried-forward method used (assumes that participants who discontinue remain stable).
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasHigh riskSponsored by manufacturer of quetiapine.

vs HLP - Velligan 2002

Methods

Allocation: randomised, no further details.
Blinding: Double-blinded, no further details.
Duration: 24 weeks.
Design: multicentre.
Setting: outpatient.

Consent: obtained.

Loss: described.

Participants

Diagnosis: Schizophrenia (DSM-III-R).
N = 116*.
Sex: male 43, female 15.
Age: mean 39.5 years.
History: conventional antipsychotic doses equivalent to 30 mg/day or less of haloperidol; in full or partial remission; BPRS no more than 3 for items assessing 'conceptual disorganisation', 'suspiciousness', hallucinatory behaviour' and 'unusual thought content'; CGI-S no more than 4.

Exclusion:physical disorder or laboratory finding that made it inappropriate to receive study medication.

Interventions

1. Quetiapine: 300 mg/day. N = 41.

2. Quetiapine: 600 mg/day. N = 52.

2. Haloperidol: 12 mg/day. N = 23.

Outcomes

Leaving the study early.

Mental state: BPRS.

Cognition: Cognitive summary score (Stroop-Color-Word, Hopkins Verbal Learning, Symbol Digit Substitution, Trials A and B, Paragraph Recall, Verbal Fluency, Pattern Memory and Paced Serial Addition.)

Notes

*The participants were a sub-sample form 301 participants and only 58 participants were included in the analysis.

BPRS = Brief Psychiatric Rating Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble blind, no further details.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
High riskData for leaving study early were available but total was high (50%).
ITT was not used.
Selective reporting (reporting bias)Low riskEvidence of selective reporting was not found.
Other biasHigh riskSponsored by manufacturers of quetiapine.

vs HLP - Zou 2007

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 8 weeks.
Setting: inpatients, Leshan mental health hospital, Sichuang, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 72.
Age: 18-60 years.
Sex: male and female.
History: duration ill mean 4.2 years, SD 2 years.
Exclusion: severe physical illness, delayed/impaired mental development, pregnant, abnormal lab test results, received depot antipsychotics within 1 month prior to the trial.
Interventions1. Quetiapine: 400-800 mg/day. N = 36.
2. Haloperidol: 8-32 mg/day. N = 36.
Outcomes

Mental state: PANSS scores.

Global state: no clinical improvement*.

Adverse events.

Notes

*PANSS decreased rate < 25%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS scores, and some of the PANSS outcomes measured, but not reported, even though these are subscale scores.
Other biasLow riskNone obvious.

vs PERPHEN - Chen 2007a

MethodsAllocation: randomised, no further details.
Blinding: not reported.
Duration: 10 weeks.
Setting: inpatients and outpatients, Chaoyang district mental health centre, Beijing, China.
Participants

Diagnosis: schizophrenia (CCMD-3).

N = 61.
Age: > 60years.
Sex: male and female.
History: duration ill: 15-30 years.
Exclusion: severe physical illness.

Interventions1. Quetiapine: 422.6 ± 108.4 mg/day. N = 31.
2. Perphenazine: 21.0 ± 16.4 mg/day. N = 30.
Outcomes

Mental state: PANSS.

Global state: No clinical improvement*.
Adverse events.

Unable to use: medication compliance, no n number reported.

Notes

*PANSS decreased rate <40%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS was measured, but not reported.
Other biasLow riskNone obvious.

vs PERPHEN - L'rman 2005

Methods

Allocation: randomised, no further details.
Blinding: double, identical-appearing capsules.
Duration: 18 months.
Design: parallel, multicentre (57 US. sites).
Setting: inpatient and outpatient, USA.

Consent: obtained.

Loss: described.

Participants

Diagnosis: Schizophrenia (DSM-IV).
N = 1,493**.
Sex: male 1080, female 380.
Age: 18-65 years (mean Quetiapine = 40.9 years, mean Perphenazine = 40.0 years).
History: duration ill, age of onset: not reported.

Exclusion: diagnosis of schizoaffective disorder, mental retardation, or other cognitive disorders; history of serious adverse reactions to the proposed treatments; one schizophrenic episode; history of treatment resistance; pregnant or breastfeeding; serious and unstable medical condition.

Interventions

1. Quetiapine: range 200-800 mg/day, mean 543.4 mg/day. N = 337.

2. Perphenazine: range 8-32 mg/day, mean 20.8 mg/day. N = 261.

3. Olanzapine: range 7.5-30 mg/day, mean 20.1 mg/day. N = 336.

4. Riperidone: range 1.5-6.0 mg/day, mean 3.9 mg/day. N = 341.

5. Ziprasidone*: range 40-160 mg/day, mean 112.8 mg/day. N = 185.

Outcomes

Leaving the study early: any cause, lack of efficacy, intolerability, patient's decision.
Global state: CGI.
Mental state: PANSS (total, positive, negative sub-score).

Cognitive***: Compostite score (processing speed, reasoning, working memory, verbal memory, vigilance)
Adverse effects: any serious adverse event, suicidal attempt, suicidal ideation, extrapyramidal side effects (AIMS, BAS, SAS), weight change, prolactin level.

Notes

* Ziprasidone was included after its approval by the US FDA.

**Only 1,460 participants were randomised.

*** The composite score consisted of five neurocognitive domains scores which were calculated from controlled oral word, category instances, grooved pegboard, Wechsler adult intelligence-revised, Wisconsin card sorting test, Wechsler intelligence scale for children-third edition mazes, Hopkins verbal learning test, computerized test of visuospatial working memory, letter-number sequencing test and continuous performance test.

PANSS = Positive and Negative Syndrome Scale.

CGI = Clinical Global Impression.

AIMS = Abnormal Involuntary Movement Scale.

BAS = Barnes Akathisia Scale.

SAS = Simpson Angus scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskDouble blind, identical-appearing capsules.
Whether blinding was successful was not examined.
Incomplete outcome data (attrition bias)
All outcomes
Low riskData for leaving study early were available but total was high (74%).
Last-observation-carried-forward method used (assumes that participants who discontinue remain stable).
Selective reporting (reporting bias)Low riskEvidence for selective reporting was not found.
Other biasUnclear risk

Dose ranges of each antipsychotic drugs were quite different.

Allocation to ziprasidone treatment was not possible from the start.

vs PERPHEN - Yi 2006

  1. a

    BMI = body mass index
    ECT= electroconvulsive therapy
    ITT = intention-to-treat
    SD = standard deviation

MethodsAllocation: randomised, no further details.
Blinding: not stated.
Duration: 12 weeks.
Setting: community and hospital, Sichuang, China.
ParticipantsDiagnosis: schizophrenia (CCMD-3).
N = 81.
Age: mean˜11.4 years, SD˜3 years.
Sex: male and female.
History: duration ill 1-14months.
Exclusion: severe physical illness, abnormal lab test results.
Interventions1. Quetiapine: 560 ± 50 mg/day. N = 41.
2. Perphenazine: 7.55 ± 3.28 mg/day. N = 40.
Outcomes

Mental state: PANSS scores.

Global state: no clinical improvement*.

Adverse events.

Notes

*PANSS decreased rate =/<25%.

CCMD = Chinese Classification of Mental Disorders.

PANSS = Positive and Negative Syndrome Scale.

TESS = Treatment Emergent Symptom Scale.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskRandomised, no further details.
Allocation concealment (selection bias)Unclear riskNo details.
Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo incomplete data.
Selective reporting (reporting bias)High riskTESS scores measured, but not reported.
Other biasLow riskNone obvious.

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
  1. a

    CCMD-3: Chinese Classification of Mental Disorders, third version
    GLU: glucose
    PANSS: Positive and Negative Syndrome Scale
    RCT: randomised controlled trial
    TC: total cholesterol
    TG: triglycerides
    vs: versus
    WCST: Wisconsin Card Sorting Test

Buckley 2001

Allocation: randomised.

Participants: schizophrenia.

Intervention: quetiapine vs haloperidol.

Outcome: no usable data.

Cheng 2003

Allocation: randomised.

Participants: schizophrenia.

Intervention: quetiapine + routine antipsychotics vs chlorpromazine + routine antipsychotics.

Du 2004

Allocation: quasi randomisation - according to hospital admission order.

Participants: schizophrenia.

Interventions: quetiapine vs chlorpromazine.

Glick 2005

Allocation: randomised.

Participants: schizophrenia.

Intervention: quetiapine vs haloperidol decanoate.

Incomplete outcome data: data for leaving study early were available but total was high (37%) and Intention-to-treat was not used.

Sponsored by manufacturers of quetiapine.

Goldstein 1999

Allocation: randomised.

Participants: schizophrenia.

Intervention: quetiapine vs haloperidol.

Outcome: no usable data.

Grecu 2006

Allocation: randomised.
Participants: schizophrenia.

Intervention: quetiapine vs haloperidol.
Outcome: no usable data (data reported in graph).

Guo 2006Allocation: randomised.
Participants: schizophrenia.
Intervention: quetiapine vs chlorpromazine.
Outcome: no usable data (only blood sugar level, TC, TG, GLU were reported).
Hammond 2001

Allocation: unclear.
Participants: schizophrenia, schizoaffective disorder, schizophreniform disorder.

Intervention: quetiapine vs haloperidol.
Outcome: no usable data.

Jones 2006Allocation: randomised.
Participants: schizophrenia, schizoaffective disorder or delusional disorder.
Interventions: first-generation antipsychotic drugs vs second-generation antipsychotic drugs as group, no usable data.
Keleman 2006

Allocation: unclear.
Participants: schizophrenia.

Intervention: quetiapine vs haloperidol.
Outcome: no usable data.

Kelly 2005

Allocation: randomised.
Participants: schizophrenia.

Intervention: quetiapine vs fluphenazine.
Outcome: no usable data (only thyroid functions were reported).

Kelly 2006

Allocation: randomised.
Participants: schizophrenia.

Intervention: quetiapine vs fluphenazine.
Outcome: no usable data (data reported in graph).

Kong 2003Allocation: randomised.
Participants: schizophrenia (CCMD-3).
Intervention: quetiapine vs chlorpromazine.
Outcome: no usable data (only blood serum concentration was reported).
Lee 2001Post-hoc analysis of pooled data from 3 RCT's.
Loza 2001

Allocation: randomised.
Participants: schizophrenia.

Intervention: quetiapine vs perphenazine.
Outcome: abstract with no usable data.

Ma 2004Allocation: controlled clinical trial, not randomised.
Nai 2007Allocation: randomised.
Participants: schizophrenia.
Intervention: quetiapine vs chlorpromazine.
Outcome: no usable data (WCST subscale scores).
Qiu 2003Allocation: controlled clinical trial, not randomised.
Participants: schizophrenia.
Intervention: quetiapine vs haloperidol.
Outcome: PANSS.
Reveley 2001

Allocation: randomised.
Participants: schizophrenia.

Intervention: quetiapine vs chlorpromazine.
Outcome: not available.

Richardson 2005

Allocation: randomised.
Participants: schizophrenia.

Intervention: quetiapine vs chlorpromazine.
Outcome: abstract with no usable data.

Sharma 2002

Allocation: randomised.
Participants: schizophrenia.

Intervention: quetiapine vs haloperidol.
Outcome: not available.

Zhang 2007Allocation: quasi randomisation - randomised according to hospital admission number.

Ancillary