Fluvoxamine versus other anti-depressive agents for depression

  • Review
  • Intervention

Authors


Abstract

Background

Fluvoxamine, one of the oldest selective serotonin reuptake inhibitors (SSRIs), is prescribed to patients with major depression in many countries. Several studies have previously reviewed the efficacy and tolerability of fluvoxamine for the treatment of major depression. However, these reviews are now outdated.

Objectives

Our objective is to evaluate the effectiveness, tolerability and side effect profile of fluvoxamine for major depression in comparison with other anti-depressive agents, including tricyclics (TCAs), heterocyclics, other SSRIs, SNRIs, other newer agents and other conventional psychotropic drugs.

Search methods

We searched the Cochrane Collaboration Depression, Anxiety and Neurosis Controlled Trials Register. Trial databases and ongoing trial registers in North America, Europe, Japan and Australia, were handsearched for randomised controlled trials. We checked reference lists of the articles included in the review, previous systematic reviews and major textbooks of affective disorder for published reports and citations of unpublished research. The date of last search was 31 August 2008.

Selection criteria

We included all randomised controlled trials, published in any language, that compared fluvoxamine with any other active antidepressants in the acute phase treatment of major depression.

Data collection and analysis

Two independent review authors inspected citations and abstracts, obtained papers, extracted data and assessed the risk of bias of included studies. We analysed dichotomous data using odds ratios (ORs) and continuous data using the standardised mean difference (SMD). A random effects model was used to combine studies.

Main results

A total of 54 randomised controlled trials (n = 5122) were included. No strong evidence was found to indicate that fluvoxamine was either superior or inferior to other antidepressants regarding response, remission and tolerability. However, differing side effect profiles were evident, especially with regard to gastrointestinal side effects of fluvoxamine when compared to other antidepressants. For example, fluvoxamine was generally associated with a higher incidence of vomiting/nausea (versus imipramine, OR 2.23, CI 1.59 to 3.14; versus clomipramine, OR 2.13, CI 1.06 to 4.27; versus amitriptyline, OR 2.86, CI 1.31 to 2.63).

Authors' conclusions

We found no strong evidence that fluvoxamine was either superior or inferior to any other antidepressants in terms of efficacy and tolerability in the acute phase treatment of depression. However, differing side effect profiles were evident. Based on these findings, we conclude that clinicians should focus on practical or clinically relevant considerations, including these differences in side effect profiles.

Résumé scientifique

Fluvoxamine versus autres agents antidépresseurs pour le traitement de la dépression

Contexte

La fluvoxamine, l'un des plus anciens inhibiteurs sélectifs de la recapture de la sérotonine (ISRS), est prescrite aux patients atteints de dépression majeure dans de nombreux pays. Plusieurs études ont examiné l'efficacité et la tolérance de la fluvoxamine dans le traitement de la dépression majeure. Néanmoins, ces revues sont aujourd'hui dépassées.

Objectifs

Notre objectif est d'évaluer l'efficacité, la tolérance et le profil d'effets secondaires de la fluvoxamine dans la dépression majeure par rapport à d'autres agents antidépresseurs, y compris les tricycliques (ATC), les hétérocycliques, d'autres ISRS, des IRSN, d'autres agents plus récents et d'autres psychotropes traditionnels.

Stratégie de recherche documentaire

Nous avons consulté le registre des essais contrôlés de la Collaboration Cochrane sur la dépression, l'anxiété et la névrose. Nous avons effectué une recherche manuelle dans des bases de données d'essais cliniques et des registre d’essais en cours d'Amérique du Nord, d'Europe, du Japon et d'Australie afin d'identifier des essais contrôlés randomisés. Nous avons examiné les références bibliographiques des articles inclus dans cette revue, des précédentes revues systématiques et des ouvrages de référence sur les troubles affectifs afin d'identifier des rapports publiés et des références à des recherches non publiées. La dernière recherche a été effectuée le 31 août 2008.

Critères de sélection

Nous avons inclus tous les essais contrôlés randomisés, sans restriction de langue, comparant de la fluvoxamine à n'importe quel autre antidépresseur actif dans le traitement de la phase aiguë de la dépression majeure.

Recueil et analyse des données

Deux auteurs de revue indépendants ont examiné les références bibliographiques et les résumés, obtenu les articles, extrait les données et évalué le risque de biais des études incluses. Nous avons analysé les données dichotomiques à l'aide des rapports des cotes, et les données continues à l'aide de la différence moyenne standardisée (DMS). Un modèle à effets aléatoires a été utilisé pour combiner les études.

Résultats principaux

Au total, 54 essais contrôlés randomisés (n = 5 122) ont été inclus. Aucune preuve solide n'indique que la fluvoxamine est supérieure ou inférieure à d'autres antidépresseurs en termes de réponse, de rémission et de tolérance. Néanmoins, différents profils d'effets secondaires étaient identifiés, en particulier concernant les effets secondaires gastro-intestinaux de la fluvoxamine par rapport à ceux d'autres antidépresseurs. À titre d'exemple, la fluvoxamine était généralement associée à une incidence supérieure de vomissements/nausées (par rapport à l'imipramine, rapport des cotes de 2,23, IC entre 1,59 et 3,14 ; par rapport à la clomipramine, rapport des cotes de 2,13, IC entre 1,06 et 4,27 ; par rapport à l'amitriptyline, rapport des cotes de 2,86, IC entre 1,31 et 2,63).

Conclusions des auteurs

Nous n'avons identifié aucune preuve solide de la supériorité ou de l'infériorité de la fluvoxamine par rapport à n'importe quel autre antidépresseur en termes d'efficacité et de tolérance dans le traitement de la phase aiguë de la dépression. Néanmoins, différents profils d'effets secondaires étaient identifiés. Sur la base de ces constatations, nous concluons que les cliniciens devraient se concentrer sur des considérations pratiques ou cliniquement pertinentes, notamment les différents profils d'effets secondaires.

Resumo

Fluvoxamina versus outros agentes antidepressivos no tratamento da depressão

Introdução

A fluvoxamina, um dos mais antigos inibidores seletivos de recaptação de serotonina (ISRSs), é prescrita para pacientes com depressão maior em muitos países. Vários estudos já revisaram a eficácia e a tolerabilidade da fluvoxamina no tratamento da depressão maior. Entretanto, essas revisões estão desatualizadas.

Objetivos

O nosso objetivo é avaliar a efetividade, a tolerabilidade e o perfil de efeitos colaterais da fluvoxamina no tratamento da depressão maior em comparação com outros agentes antidepressivos, incluindo os tricíclicos (ADTs), heterocíclicos, outros ISRSs, inibidores da recaptação de serotonina-noradrenalina (IRSNs), outros agentes mais novos e outras drogas psicotrópicas convencionais.

Métodos de busca

Pesquisamos nas bases Cochrane Collaboration Depression, Anxiety and Neurosis Controlled Trials Register. Fizemos busca manual nos registros de ensaios clínicos e de estudos em andamento na América do Norte, Europa, Japão e Austrália para encontrar ensaios clínicos randomizados. Fizemos também uma busca por citações relevantes de estudos publicados ou não nas listas de referências dos artigos incluídos na revisão, em revisões sistemáticas anteriores e em livros textos sobre transtornos afetivos. A data da última busca foi 31 de agosto de 2008.

Critério de seleção

Incluímos todos os ensaios clínicos randomizados, publicados em qualquer língua, que compararam a fluvoxamina com quaisquer outros antidepressivos no tratamento da fase aguda da depressão maior.

Coleta dos dados e análises

Dois autores examinaram independentemente as citações e os resumos, obtiveram os artigos, extraíram os dados e avaliaram o risco de viés dos estudos incluídos. Analisamos os dados dicotômicos usando odds ratios (ORs) e, para os dados contínuos, usamos a diferença padronizada de médias (SMD). Um modelo de efeitos aleatórios foi usado para combinar os estudos.

Principais resultados

Incluímos um total de 54 ensaios clínicos randomizados (n = 5122). Não encontramos evidência robusta que indicasse que a fluvoxamina fosse superior ou inferior aos outros antidepressivos com relação a resposta, remissão e tolerabilidade. Contudo, as diferenças nos perfis de efeitos colaterais foram evidentes, especialmente no tocante aos efeitos colaterais gastrointestinais da fluvoxamina quando comparada aos outros antidepressivos. Por exemplo, a fluvoxamina foi associada a maior incidência de vômitos/náuseas (versus imipramina, OR 2,23, 95% CI 1,59 à 3,14; versus clomipramina, OR 2,13, 95% CI 1,06 to 4,27; versus amitriptilina, OR 2.86, 95% CI 1,31 to 2,63).

Conclusão dos autores

Não encontramos evidência robusta de que a fluvoxamina fosse superior ou inferior a quaisquer outros antidepressivos em termos de eficácia e tolerabilidade na fase aguda do tratamento da depressão. Todavia, as diferenças nos perfis de efeitos colaterais foram evidentes. Com base nesses achados, concluímos que os médicos devem focar em considerações de ordem prática ou clinicamente relevantes, incluindo estas diferenças nos perfis de efeitos colaterais.

Notas de tradução

Tradução do Centro Cochrane do Brasil (Bárbara Perdigão Stumpf)

Plain language summary

Fluvoxamine versus other anti-depressive agents for depression

Major depression is a severe mental illness characterised by a persistent and unreactive low mood and loss of all interest and pleasure, usually accompanied by a range of symptoms including appetite change, sleep disturbance, fatigue, loss of energy, poor concentration, psychomotor symptoms, inappropriate guilt and morbid thoughts of death. Although pharmacological and psychological interventions are both effective for major depression, antidepressant (AD) drugs remain the mainstay for treatment of moderate or severe depression. Fluvoxamine is one of the oldest selective serotonin reuptake inhibitors (SSRIs) and is prescribed to patients with major depression in many countries. This review reports trials comparing fluvoxamine with other antidepressants for treatment of major depression. We found no strong evidence that fluvoxamine was either superior or inferior to any other antidepressants in terms of efficacy and tolerability in the acute phase treatment of depression. However, there is evidence of differing side-effect profiles, especially when comparing gastrointestinal side effects between fluvoxamine and tricyclic antidepressants (TCAs). Based on these findings, we conclude that clinicians should focus on practical or clinically relevant considerations including these differences in side effect profiles.

Résumé simplifié

Fluvoxamine versus autres agents antidépresseurs pour le traitement de la dépression

La dépression majeure est une maladie mentale grave, caractérisée par une humeur morose persistante et non réactive et la perte de tout intérêt et plaisir, le plus souvent accompagnée d'un éventail de symptômes, y compris le changement d'appétit, les troubles du sommeil, la fatigue, la perte d'énergie, le manque de concentration, des symptômes psychomoteurs, des pensées inappropriées de culpabilité et morbides de mort. Bien que les interventions pharmacologiques et psychologiques soient toutes deux efficaces dans la dépression majeure, les antidépresseurs (AD) demeurent le pilier du traitement de la dépression modérée à sévère. La fluvoxamine, l'un des plus anciens inhibiteurs sélectifs de la recapture de la sérotonine (ISRS), est prescrite aux patients atteints de dépression majeure dans de nombreux pays. Cette revue examine les essais comparant de la fluvoxamine à d'autres antidépresseurs dans le traitement de la dépression majeure. Nous n'avons identifié aucune preuve solide de la supériorité ou de l'infériorité de la fluvoxamine par rapport à n'importe quel autre antidépresseur en termes d'efficacité et de tolérance dans le traitement de la phase aiguë de la dépression. Néanmoins, certaines preuves suggèrent qu'il existe différents profils d'effets secondaires, en particulier lorsque l'on compare les effets secondaires gastro-intestinaux de la fluvoxamine à ceux des antidépresseurs tricycliques (ATC). Sur la base de ces constatations, nous concluons que les cliniciens devraient se concentrer sur des considérations pratiques ou cliniquement pertinentes, notamment sur les différents profils d'effets secondaires.

Notes de traduction

Traduit par: French Cochrane Centre 1st October, 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.

Resumo para leigos

Fluvoxamina versus outros agentes antidepressivos no tratamento da depressão

A depressão maior é uma doença mental grave, caracterizada por rebaixamento de humor persistente e não reativo, além da perda de interesse e prazer. É geralmente acompanhada de uma gama de sintomas, incluindo mudança do apetite, distúrbio do sono, fadiga, falta de energia, dificuldade de concentração, sintomas psicomotores, culpa indevida e pensamentos mórbidos de morte. Embora tanto as intervenções farmacológicas como as psicológicas sejam efetivas para o tratamento da depressão maior, os antidepressivos (ADs) continuam sendo os pilares do tratamento da depressão moderada ou grave. A fluvoxamina é um dos inibidores seletivos de recaptação de serotonina (ISRSs) mais antigos e é prescrita para pacientes com depressão maior em muitos países. Esta revisão mostra os resultados de estudos que compararam a fluvoxamina com outros antidepressivos no tratamento da depressão maior. Não encontramos evidência robusta de que a fluvoxamina fosse superior ou inferior a quaisquer outros antidepressivos em termos de eficácia e tolerabilidade na fase aguda do tratamento da depressão. Entretanto, há evidência de diferenças nos perfis de efeitos colaterais, especialmente quando se comparam os efeitos colaterais gastrointestinais da fluvoxamina com os dos antidepressivos tricíclicos (ADTs). Com base nesses achados, concluímos que os clínicos devem focar em considerações de ordem prática ou clinicamente relevantes, incluindo essas diferenças nos perfis de efeitos colaterais.

Notas de tradução

Tradução do Centro Cochrane do Brasil (Bárbara Perdigão Stumpf)

Background

Description of the condition

Major depression is generally diagnosed when a persistent and unreactive low mood and loss of all interest and pleasure are usually accompanied by a range of symptoms including appetite change, sleep disturbance, fatigue, loss of energy, poor concentration, psychomotor symptoms, inappropriate guilt, and morbid thoughts of death (APA 1994). In 2002, major depression was the third leading health burden in the world, following only lower respiratory infections and HIV/AIDS, and accounting for 4.5% of total human suffering related to health concerns. Moreover, the incidence of depression is expected to rise during the next 20 years (WHO 2006). The depressed condition is associated with a marked personal, social and economic morbidity, coupled with a loss of functioning and productivity, which creates significant demands on health service provider workloads (NICE 2004). In the USA, Greenberg 2003 estimated the economic burden of depression to be just over $83 billion in 2000. Of this total, $26 billion came from direct treatment costs, $5 billion came from suicide-related costs, and $52 billion came from workplace costs. These figures were also suspected to underestimate the true economic burden of the disease, as they did not take into account factors such as the burden on family members and caregivers, the cost of lost productivity while at work, and costs associated with those who remain untreated (Greenberg 2005).

Description of the intervention

Fluvoxamine ((E)-5-methoxy-1-[4-(trifluoromethyl)phenyl]pentan-1-one O-2-aminoethyl oxime) is a selective serotonin reuptake inhibitor (SSRI) that has been available as an antidepressant since 1983 in many countries – 87 countries and regions as of 2006, including some European countries and Japan. It is also available in many countries for anxiety disorders, including obsessive-compulsive disorder and social anxiety disorder. Fluvoxamine is structurally different from the tricyclic antidepressants (TCAs), heterocyclics, and other classes of antidepressants, and also differs chemically from various other SSRIs. For example, fluvoxamine is the only monocyclic SSRI and it belongs to the 2-aminoethyloximethers of aralkylketones (Claassen 1977; Fuller 1987). Therefore, some differential clinical potency may be expected, not only between the drugs classes but also among the SSRIs.

How the intervention might work

Fluvoxamine is well absorbed after oral administration and is widely distributed throughout the body. Plasma protein binding of fluvoxamine is low (77%), compared with other SSRIs. Fluvoxamine displays nonlinear, steady-state pharmacokinetics throughout the therapeutic range, with disproportionately higher plasma concentrations at higher doses (Perucca 1994). However, plasma fluvoxamine concentrations show no clear relationship with patient responses to the antidepressant or to the severity of adverse effects. Fluvoxamine pharmacokinetics remains unaltered by increasing age or by renal impairment. Fluvoxamine is metabolized in the liver by the cytochrome P450 (CYP) enzyme system. It has a prominent affinity for the CYP1A2 isozyme, a lesser affinity for the CYP3A4 and CYP2C isozymes, and a minimal affinity for CYP2D6. Fluvoxamine impairs metabolic elimination of a number of drugs, including TCAs (tertiary, but not secondary, amines), alprazolam, bromazepam, diazepam, theophylline, propranolol and, possibly, carbamazepine. It generates no active metabolites. Smoking is known to increase CYP1A2 activity, and smokers appear to have lower serum concentration of fluvoxamine compared with non-smokers (Spigset 1995). The drug is eliminated with a mean half-life of 15 hours, with a range from nine to 28 hours. Excretion is primarily in the urine, predominantly as metabolites (van Harten 1995).

Why it is important to do this review

Although pharmacological and psychological interventions are both effective for major depression, antidepressant (AD) drugs remain the mainstay for treatment of moderate or severe depression (APA 2000; Ellis 2004; NICE 2004). Many different AD agents are available, including TCAs, monoamine oxidase inhibitors (MAOIs), selective serotonin reuptake inhibitors (SSRIs), serotonin-noradrenaline reuptake inhibitors (SNRIs: venlafaxine, duloxetine, milnacipran), and other newer agents (mirtazapine, reboxetine, bupropion). In many Western countries, AD consumption has risen dramatically over the last 20 years, mainly because of the increasing consumption of SSRIs and newer ADs, which have progressively become the most commonly prescribed ADs (Ciuna 2004; Guaiana 2005). SSRIs are generally better tolerated than TCAs (Barbui 2000), and there is evidence of similar efficacy (Anderson 2000; Geddes 2000; Williams 2000). However, head-to-head comparisons have provided contrasting findings. Amitriptyline, for example, may have an edge over SSRIs in terms of efficacy (Guaiana 2007), while individual SSRIs and SNRIs may differ in terms of efficacy and tolerability (Cipriani 2005; Cipriani 2009; Puech 1997; Smith 2002; ).

Two systematic reviews on fluvoxamine exist in current literature. Burton (Burton 1991) reviewed 17 double-blind comparative studies between fluvoxamine and other ADs in depressed patients. Ware (Ware 1997) reviewed 31 controlled trials of fluvoxamine in the pharmacotherapy of depression. These reviews are now quite outdated and neither has provided meta-analytic summaries.

A group of researchers therefore agreed to join forces under the rubric of the Meta-Analyses of New Generation Antidepressants Study Group (MANGA Study Group), in order to systematically review all available evidence for each specific newer antidepressant. As of February 2010, we have completed an individual review for fluoxetine (Cipriani 2006), mirtazapine (Watanabe 2008), milnacipran (Nakagawa 2007), escitalopram (Cipriani 2009a) and sertraline (Cipriani 2009b) and have published the protocols for citalopram (Imperadore 2007), duloxetine (Nose 2007), paroxetine (Cipriani 2007b) , venlafaxine (Cipriani 2007c) and reboxetine (Churchill 2009). The multiple-treatment meta-analysis of 12 new-generation antidepressants has also been published (Cipriani 2009).

In the present review, we report head-to-head comparisons not only between fluvoxamine and the other 11 new-generation antidepressants, but also between fluvoxamine and older antidepressants, providing detailed accounts of their comparative side effect profiles.

Objectives

1. To determine the efficacy of fluvoxamine compared to other anti-depressive agents, including older antidepressants such as TCAs and newer ones such as SSRIs, in alleviating the acute symptoms of depression.
2. To review acceptability of treatment with fluvoxamine compared with that of other antidepressive agents.
3. To investigate the adverse effects of fluvoxamine compared to other antidepressive agents.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials (RCTs) were included. Quasi-randomised trials, such as those allocating by using alternate days of the week, were excluded. For trials that have a crossover design, only results from the first randomisation period were considered.

Types of participants

The review included patients 18 or older, of both sexes, with a primary diagnosis of depression and studies adopting standardised criteria (DSM-III / DSM- III-R, DSM-IV (APA 2000), ICD-10 (WHO 1992), Feighner criteria (Feighner 1972) or Research Diagnostic Criteria (Spitzer 1972) to define patients suffering from unipolar major depression. Studies using ICD-9 were excluded, as it has only disease names and no diagnostic criteria.

We included the following subtypes of depression: chronic, with catatonic features, with melancholic features, with atypical features, with postpartum onset, and with seasonal pattern. We also included studies in which up to 20% patients presented depressive episodes in bipolar affective disorder. When depressive patients in the trial had psychotic features, we included those studies in which up to 20% patients presented psychotic features. A concurrent secondary diagnosis of another psychiatric disorder was not considered an exclusion criterion. A concurrent primary diagnosis of Axis I or II disorders was an exclusion criterion. AD trials in depressive patients with a serious concomitant medical illness were excluded.

Types of interventions

We have examined fluvoxamine intervention in comparison with conventional treatment of acute depression. We also examined fluvoxamine intervention in comparison with non-conventional (e.g., herbal products, such as Hypericum) anti-depressive agents (Linde 2008). Trials in which fluvoxamine was compared to another type of psychopharmacological agent (i.e., anxiolytics, anticonvulsants, antipsychotics or mood-stabilizers) were excluded. We also excluded trials in which fluvoxamine was used as an augmentation strategy.

Eligible intervention:

1. Fluvoxamine: any dose and mode or pattern of administration.

Eligible comparators:

2. Conventional anti-depressive agents: any dose and mode or pattern of administration.
2.1 Tricyclics (TCAs)
2.2 Heterocyclics
2.3 SSRIs
2.4 SNRIs
2.5 MAOIs or newer antidepressants (ADs)
2.6 Other conventional psychotropic drugs

3. Non-conventional anti-depressive agents
3.1 Herbal products
3.2 Other non-conventional anti-depressive agents

Types of outcome measures

Primary outcomes
1. Response - acute phase

We examined cases regarding the number of patients (1) who responded to treatment by showing a reduction of at least 50% on the Hamilton Rating Scale for depression (HRSD) (Hamilton 1960), Montgomery Åsberg Depression Rating Scale (MADRS) (Montgomery 1979), or any other depression scale, depending on the study authors' definition or (2) who were "much or very much improved" (score 1 or 2) on the CGI-Improvement scale (Guy 1976) out of the total number of randomised patients. Where both are provided, we prefer the former criteria for judging response. The original authors' definitions of response and remission were not used in this review, to avoid possible outcome reporting bias (Furukawa 2007).

When studies report response rates at various time points throughout the trial, we have determined a priori to subdivide the treatment indices - since one systematic review suggested that SSRIs begin to have observable beneficial effects in depression during the first week of treatment - as follows (Taylor 2006):

(i) Response - early phase: between 1 and 4 weeks, with the time point closest to 2 weeks given preference.
(ii) Response - acute phase: between 6 and 12 weeks, with preference given to the time point given in the original study as the study endpoint.
(iii) Response - follow-up phase: between 4 and 6 months, with the time point closest to 24 weeks given preference.

The acute phase treatment response rates were our primary outcome of interest.

Secondary outcomes
1. Response - early phase, and follow-up phase
2. Remission - early phase, acute phase, and follow-up phase

We are interested in the number of patients who achieved remission, (1) showing =<7 on HRSD-17, =<8 on for all the other longer versions of HRSD, and =<11 on MADRS or (2) who were "not ill or borderline mentally ill" (score 1 or 2) on the CGI-Severity score out of the total number of randomised patients. Where both were provided, we preferred the former criterion for judging remission.

3. Group mean scores at the end of the trial and change score on depression scale
4. Social adjustment, social functioning, including the Global Assessment of Function (GAF) scores

(Hall 1995)

5. Health-related quality of life (QOL)

We limited ourselves to SF-12 (Ware 1998); SF-36 (Ware 1992), HoNOS (Wing 1998) and the WHO 2009-QOL (WHOQOL Group 1998).

6. Costs to health care services.
7. Tolerability
7.1 Total dropout

Number of patients who dropped out during the trial as a proportion of the total number of randomised patients.

7.2 Dropout due to inefficacy

Number of patients who dropped out during the trial because the fluvoxamine was ineffective as a proportion of the total number of randomised patients.

7.3 Dropout due to side effects

Number of patients who dropped out during the trial due to side effects, as a proportion of the total number of randomised patients.

7.4 Number of patients experiencing at least one side effect
7.5 Number of patients experiencing the following specific side effects was sought:
- sleepiness/drowsiness
- insomnia
- dry mouth
- constipation
- problems urinating
- hypotension
- agitation/anxiety
- suicide wishes/gestures/attempts
- completed suicide
- vomiting/nausea
- diarrhoea

To avoid missing any relatively rare or unexpected side effects in the data extraction phase, we collected all side effect data reported in the literature and discussed ways to summarize them post hoc. Descriptive data regarding side effect profiles were extracted from all available studies. Only studies reporting the number of patients experiencing individual side effects were retained. Due to a lack of consistent reporting of side effects, which came primarily from the study authors' descriptions, we combined terms describing similar side effects; for example, we combined "dry mouth", "reduced salivation" and "thirst" into "dry mouth". All side effect categories were then grouped by organ system, such as neuropsychiatric, gastrointestinal, respiratory, sensory, genitourinary, dermatological and cardiovascular, in accordance with the advice of a previous study (Mottram 2006).

Search methods for identification of studies

Electronic searches

We initially identified RCTs on June 2, 2006 by following the Cochrane Collaboration Depression, Anxiety and Neurosis (CCDAN) criteria for search strategy using the register of CCDAN Review Group Controlled Trials Registers (CCDANCTR-Studies and CCDANCTR-References). The registers are compiled from systematic and regularly updated searches of the Cochrane Central Register of Controlled Trials (CENTRAL) - the most comprehensive source of reports of RCTs - , MEDLINE, EMBASE, CINAHL, PsycINFO, PSYINDEX, and LILACS and handsearched of major psychiatric and medical journals as well as conference proceedings. Trial databases (e.g., the Medicines and Healthcare products Regulatory Agency in the UK) and ongoing trial registers (e.g., clinicaltrials.gov in the USA) in North America, Europe, Japan and Australia, were handsearched for published, unpublished and ongoing RCTs.

CCDANCTR-Studies was searched using the following search strategy:
Diagnosis = Depress* or Dysthymi* or "Adjustment Disorder*" or "Mood Disorder*" or "Affective Disorder" or "Affective Symptoms"
and
Intervention = Fluvoxamine

CCDANCTR-References was searched using the following search strategy:
Keyword = Depress* or Dysthymi* or "Adjustment Disorder*" or "Mood Disorder*" or "Affective Disorder" or "Affective Symptoms"
and
Free-Text = Fluvoxamine

The researchers conducted additional searches on the CCDAN Review Group Controlled Trials Registers, MEDLINE and checked various meta-analysis and review articles on the 26th October 2009 (CCDAN Registers up-to-date as of 31 August 2008).

Searching other resources

1. Handsearches

We searched trial databases of the following drug-approving agencies for published, unpublished and ongoing controlled trials: the Food and Drug Administration (FDA) in the USA, the Medicines and Healthcare products Regulatory Agency (MHRA) in the UK, the European Medicines Agency (EMEA) in the EU, the Pharmaceuticals and Medical Devices Agency (PMDA) in Japan and the Therapeutic Goods Administration (TGA) in Australia. We also searched ongoing trial registers such as clinicaltrials.gov in the USA, International Standard Randomised Controlled Trial Number Register (ISRCTN) and the National Research Register in the UK, Nederland's Trial Register in the Netherlands, European Union Drug Regulating Authorities Clinical Trials (EudraCT) in the EU, UMIN-CTR in Japan and the Australian Clinical Trials Registry in Australia. These searches were undertaken in November 2007.

Appropriate journals and conference proceedings relating to fluvoxamine treatment for depression have already been handsearched and incorporated into the CCDANCTR databases.

2. Personal communication

Pharmaceutical companies and experts in this field were asked if they knew of any study that met the inclusion criteria of this review.

3. Reference checking

Reference lists of the included studies, previous systematic reviews and major textbooks of affective disorder written in English were checked for published reports and citations of unpublished research.

Data collection and analysis

Selection of studies

HMG and another independent review author checked to ensure that studies relating to fluvoxamine generated by the search strategies of the CCDANCTR-References and the other complementary searches met the rough inclusion criteria, firstly based on the title and abstracts. All of the studies that were rated as possible candidates by either of the two review authors were added to the preliminary list, and their full texts were retrieved. TAF and IMO then assessed all of the full text articles in this preliminary list to see if they met the strict inclusion criteria. If the raters disagreed, the final rating was made by consensus with the involvement - if necessary - of another member of the review group. Non-congruence in selection of trials was reported as akappa statistic. Considerable care was taken to exclude duplicate publications.

Data extraction and management

IMO and NW extracted data from the included studies. Again, any disagreement was discussed, and decisions were documented. If necessary, we contacted authors of studies for clarification. We extracted the following data:

(i) participant characteristics (age, sex, depression diagnosis, comorbidity, depression severity, antidepressant treatment history for the index episode, study setting);
(ii) intervention details (intended dosage range, mean daily dosage actually prescribed, co-intervention if any, fluvoxamine as investigational drug or as comparator drug, sponsorship);
(iii) outcome measures of interest from the included studies.

The results were compared with those in the completed reviews of individual antidepressants in the Cochrane Library. If the trial was a three (or more)-armed trial involving a placebo arm, the data were extracted from the placebo arm as well.

Assessment of risk of bias in included studies

Risk of bias was assessed independently by two review authors (IMO and NW) using criteria described in the Cochrane Collaboration Handbook (Higgins 2008). This set of criteria is based on evidence of associations between effect overestimation and a high risk of bias in an article, such as sequence generation, allocation concealment, blinding, incomplete outcome data and selective reporting.

The categories are defined below:
YES - low risk of bias
NO - high risk of bias
UNCLEAR - uncertain risk of bias

Measures of treatment effect

All comparisons were performed between fluvoxamine and comparator ADs as a class and as individual ADs.

1. Dichotomous data

For dichotomous, or event-like, data, odds ratios (ORs) were calculated with its 95% confidence interval (CI). For statistically significant results, 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

For continuous data, mean differences (MD) or standardized mean differences (SMD) - where different measurement scales, were calculated with its 95% CI.

Unit of analysis issues

1. 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, even 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 major depression, we only used data from the first phase of the cross-over studies.

2. Cluster-randomised trials

No cluster randomised trials were identified for this version of the review. Should they be identified in a future update, we plan to use the generic inverse variance technique, if such trials have been appropriately analysed taking into account intraclass correlation coefficients to adjust for cluster effects.

Dealing with missing data

1. Dichotomous data

Responders and remitters to treatment were calculated on the strict intention-to-treat (ITT) basis: dropouts were included in this analysis. Where participants have been excluded from the trial before the endpoint, we have assumed that they experienced a negative outcome by the end of the trial (e.g., failure to respond to treatment). We examined the validity of this decision in the sensitivity analyses by applying worst- and best-case scenarios. If a statistically significant difference was found, the number needed to treat (NNT) was calculated from an odds ratio obtained by a meta-analysis (Higgins 2008). We applied the loose ITT analyses for continuous variables, whereby all the patients with at least one post-baseline measurement were represented by their last observations carried forward (LOCF), with due consideration of the potential bias and uncertainty introduced.

When dichotomous outcomes were not reported but baseline mean, endpoint mean and those standard deviation (SD) of the HRSD (or other depression scale) were provided, we converted continuous outcome data expressed as mean and SD into the number of responding and remitted patients, according to the validated imputation method (Furukawa 2005). We examined the validity of this imputation in the sensitivity analyses. Where SDs were not reported, authors were asked to supply the data. When only the standard error (SE) or t-statistics or P values are reported, SDs were calculated according to Altman (Altman 1996). In the absence of data from the authors, we substituted SDs by those reported in other studies in the review (Furukawa 2006).

2. Continuous data

When there were missing data and the method of "last observation carried forward" (LOCF) had been used to do an ITT analysis, then the LOCF data were used. When SDs were missing, we presented data descriptively.

Assessment of heterogeneity

Heterogeneity between studies was assessed by visual inspection of the results in the forest plots. Statistic (the I2statistic and the Q statistic) were interpreted with caution, since non-significant results of statistical tests for heterogeneity cannot be regarded as evidence of heterogeneity (Higgins 2008). If the CIs for the results of individual comparisons had poor overlap, I2 was equal to or more than 50% and P values were smaller than 0.1 (Higgins 2003), potential sources of heterogeneity were investigated. We performed subgroup analyses to investigate heterogeneity (see Subgroup analysis and investigation of heterogeneity).

Assessment of reporting biases

Data from included studies were entered into a funnel plot (trial effect against trial variance) to investigate small-study effects (Sterne 2000). We used the tests for funnel plot asymmetry only when there were at least 10 studies included in the meta-analysis, and results were interpreted cautiously, with visual inspection of the funnel plots (Higgins 2008). When evidence of small-study effects was identified, possible reasons for funnel plot asymmetry, including publication bias, were investigated.

Data synthesis

The primary analysis used a random effects model (odds ratio [OR]), which had the highest generalisability in our empirical examination of summary effect measures for meta-analyses (Furukawa 2002a). The robustness of this summary measure was routinely examined by checking the fixed-effect model OR and the random effects model risk ratios (RRs). Material differences between the models were reported. A p value of less than 0.05 and a 95% confidence interval (CI) were considered statistically significant.

Fixed-effect analyses were performed routinely for the continuous outcomes as well, to investigate the effect of the choice of method on the estimates. Material differences between the models were reported. Skewed data and non-quantitative data were presented descriptively. An outcome was considered skewed when the mean was smaller than twice the SD. In terms of change score, data were difficult to depict as skewed or not as the possibility existed for negative values; therefore, we entered all of the results of this outcome into meta-analysis.

Subgroup analysis and investigation of heterogeneity

We performed the following subgroup analyses for primary outcome where possible, for the following a priori reasons. Results were interpreted with caution, since multiple comparisons could lead to false positive conclusions (Oxman 1992).

1. Fluvoxamine dosing (fixed low dosage, fixed standard dosage, fixed high dosage; flexible low dosage, flexible standard dosage, flexible high dosage)

Existing evidence implies that low dosage antidepressants may be associated with better outcomes - both in terms of effectiveness and side effects - than standard or high dosage antidepressants (Bollini 1999; Furukawa 2002b). In addition, a fixed versus flexible dosing schedule may affect estimates of treatment effectiveness (Khan 2003). In the case of fluvoxamine, based on the Defined Daily Dosage (DDD) by WHO (WHO 2009), low dosage is referred to as <100, standard dosage to >=100 but <200, and high dosage to >=200 mg/day. We categorized studies by intended maximum dosage of fluvoxamine.

2. Comparator dosing (low dosage, standard dosage, and high dosage)

It is easy to imagine that people taking a comparator drug are less likely to complete a study if they are taking a high dosage of the comparator drug. We categorized studies by the intended maximum dose of fluvoxamine based on the DDD. Since WHO 2009 does not report DDD of milnacipran, we categorized these studies based on previous reports (Lecrubier 1996; Lopez-Ibor 1996; Okamura 2006), where low dosage refers to <100, standard dosage to >=100 but <150, and high dosage to >=150 mg/day.

3. Depression severity (severe major depression, moderate/mild major depression)

"Severe major depression" was defined by a threshold baseline severity score for entry of 25 or more for HRSD and 31 or more for MADRS (Dozois 2004; Muller 2003).

4. Treatment settings (psychiatric inpatients, psychiatric outpatients, primary care)

Because depressive disorder in primary care has a different profile than that of psychiatric inpatients or outpatients (Suh 1997); it is possible that results obtained from either of these settings may not be applicable to the other settings (Depression Guideline Panel 1993).

5. Elderly patients (>=65 years of age), separately from other adult patients

Older people may be more vulnerable to side effects associated with antidepressants and decreased dosage is often recommended for them (Depression Guideline Panel 1993).

Because the number of a priori planned subgroup analyses now appears excessive in comparison with the identified studies, we will consider reducing the number of subgroup analyses or adjusting the level of significance to account for making multiple comparisons in the next update.

Sensitivity analysis

The following sensitivity analyses for primary outcome were planned a priori. By limiting the included studies to those with higher quality (analysis 1 to 5) or to those free from some "bias" (analysis 6 to 9), we examined whether the results changed and we intended to check for the robustness of the observed findings.

1. We excluded trials with unclear concealment of random allocation and/or unclear double blinding.

2. We excluded trials with a dropout rate greater than 20%.

3. We performed the worst-case scenario ITT: that all patients in the experimental group experience the negative outcome and all those in the comparison group experience the positive outcome.

4. We performed the best-case scenario ITT: that all patients in the experimental group experienced the positive outcome and all those in the comparison group experienced the negative outcome.

5. We excluded trials for which the response rates had to be calculated based on the imputation method (Furukawa 2005) and for which the SD had to be borrowed from other trials (Furukawa 2006).

6. We examined a "wish bias" by comparing the trials where fluvoxamine was used as an investigational drug, the drug that was used as a new compound, to the trials where fluvoxamine was used as a comparator, since some evidence suggests that a new antidepressant might perform worse when used as a comparator than when used as an investigational agent (Barbui 2004).

7. We excluded trials funded by, or with at least one author affiliated with, a pharmaceutical company marketing fluvoxamine. This sensitivity analysis is particularly important in light of the recent repeated findings that funding strongly affects outcomes of research studies (Als-Nielsen 2003; Bhandari 2004; Lexchin 2003; Montgomery 2004; Perlis 2005; Procyshyn 2004) and because industry sponsorship and authorship of clinical trials have increased over the past 20 years (Buchkowsky 2004).

8. We excluded studies that included patients with bipolar depression.

9. We excluded trials studies that included patients with psychotic features.

Our routine application of random effects and fixed-effect models, as well as our secondary outcomes of remission rates and continuous severity measures, may be considered additional forms of sensitivity analyses.

If the CIs of ORs in the groups did not overlap, potential sources of heterogeneity were investigated.

Results

Description of studies

Results of the search

Initially, we identified 152 references considered relevant for our review. Of these, five trials were unpublished (Coleman 1981a; Coleman 1981b; Coleman 1983; Doogan 1981; van Beek 1981), and one trial was written in Hungarian and was not retrieved. These trials has been placed on the list of studies awaiting assessment (Faludi 1989). The remaining 146 references were retrieved for more detailed evaluation (Figure 1).

Figure 1.

Flow diagram for the trials

Additional searches conducted on the 26th October 2009 (CCDAN Registers up-to-date as of 31 August 2008) found 6 trials (Berlin 1998, Donovan 1993, Entsuah 2002b, Mallick 2003, Naito 2007, Ushiroyama 2004) which might be included in the an update of this review. We have placed these articles on the list of studies awaiting assessment.

Included studies

See: Included studies, Figure 1.
We were able to include 54 studies (56 comparisons). All studies included were randomised trials. The inter-rater reliability in selection of trials was satisfactory, with weighted kappa of 0.77.

We mailed or e-mailed or phoned the authors with known contact details to obtain extra information we sought. We had replies from the authors of 23 trials. Among them, we were able to obtain unpublished data from 15 trials including 17 comparisons (Ansseau 1991a; Ansseau 1991b; Barge-Schaapveld 1995; Cassano 1986; Dalery 2003; Hackett 1998a; Hackett 1998b; Kasper 1990; Kato 2006; Nathan 1990; Nemeroff 1995; Otsubo 2005; Rechlin 1994; Remick 1994; Rossini 2005; Schoemaker 2002; Ueda 2002).

1. Length of studies

Duration of treatment was relatively brief, with a mean of 5.5 weeks (range 2 to 10 weeks). There was one 2-week study, 20 4-week studies, 24 6-week studies, five 7-week studies, three 8-week studies and one 10-week study.

2. Setting

For 18 studies, treatment occurred in a psychiatric inpatient setting; for 21 studies, treatment occurred in a psychiatric outpatient setting; and in seven studies, treatment occurred in a combined inpatient/outpatient setting. Two studies were based on primary care settings (Barge-Schaapveld 1995; Moon 1991), and six did not specify their treatment settings.

3. Participants

All trials reported that participants suffered from major depression defined by operationalised diagnostic criteria; however, some studies included less than 20% of patients with bipolar depression (Ansseau 1991a; Ansseau 1991b; Asakura 2005; de Wilde 1983; Guy 1984; Haffmans 1996; Itil 1983; Kasper 1990; March 1990; Murasaki 1998a; Ottevanger 1995; Rossini 2005). In addition, 18 studies used diagnostic criteria such as "major depressive episode" (DSM-III or IV) , "major affective disorder" (DSM-III), "depression" (Feighner criteria), or "unipolar or bipolar disorder" (Feighner criteria), and did not exclude patients with bipolar depression. Consequently, some studies might include patients with bipolar depression who were not specifically taken into account (Ansseau 1994; Barrelet 1991; Bocksberger 1993; Bougerol 1992; Brunner 1994; Cassano 1986; Coleman 1982; Dalery 2003; Dick 1983; Gonul 1999; Harris 1991a; Moon 1991; Mullin 1988; Perez 1990; Rahman 1991; Rapaport 1996; Rota 2005; Zohar 2003). Some studies included less than 20% of depressive patients with psychotic features (Ansseau 1991a; Ansseau 1991b; Ansseau 1994; Asakura 2005; Bramanti 1988; Clerc 2001; Haffmans 1996; Kasper 1990). In 25 studies, some elderly subjects (over 65 years old) were included, but the actual number of elderly people was not reported in most trials. One trial was for elderly patients only (Bocksberger 1993), while seven studies did not include any elderly patients. One trial (Claghorn 1996) only included patients with severe depression, defined by a score higher than 25 on the HRSD-17 at baseline.

4. Study size

Two studies did not report the number of patients included (Kavoussi 1999; Rota 2005). The mean sample size for remaining studies was 93, ranging from 23 (Barge-Schaapveld 1995) to 481 (Cassano 1986). The majority of the studies (38 RCTs) recruited fewer than 100 participants.

5. Interventions

There were 30 studies comparing fluvoxamine with TCAs, five studies including heterocyclics, 10 including SSRIs, three including SNRIs, four including newer antidepressants, and one comparing fluvoxamine with sulpiride and one four-arm study comparing fluvoxamine with amitriptyline, doxepine and paroxetine (Rechlin 1994). We could not find studies comparing fluvoxamine with non-conventional anti-depressive agents such as herbal products. Regarding fluvoxamine dosing, the trials included five fixed and 43 flexible schedules, and one study did not state dosing schedule. Standard doses were used in 22 studies, and high doses in 28 studies. Four studies did not state fluvoxamine dosing.

6. Outcomes

Of the included 54 studies, five studies (Brown 1986; Gonul 1999; Kavoussi 1999; Rechlin 1994; Rota 2005) did not report efficacy data, and four studies (Kavoussi 1999; Miller 2001;Rechlin 1994; Rota 2005) did not report tolerability data that could be entered into a meta-analysis. We were unable to obtain further data because we could not contact the authors by any means, nor could we obtain extra information from these authors. The majority of the identified studies (44 studies) used the HRSD as a primary or secondary outcome measure, while a minority of studies used the MADRS and Clinical Global Impression scale (CGI).

Among the 50 studies reporting dropouts due to any reason, 42 reported dropouts due to side effects. Forty-one studies reported the number of patients experiencing individual side effects. It was unclear how these adverse effects were measured in terms of either severity or duration.

Excluded studies

See: Excluded studies.

By assessing the 146 retrieved full texts, we found 52 articles that were duplicate publications. We assessed the remaining 94 studies for their eligibility and we excluded 40 articles that did not meet our inclusion criteria. Among those excluded, 19 were not randomised trials; three did not use an operationalised criteria to diagnose major depression; three included more than 20 % of participants who suffered from bipolar disorder; four included patients with major depression with psychotic features; five included an Axis I disorder other than major depression or bipolar disorder; two did not compare fluvoxamine with other antidepressants; two had a crossover design and clinical data for the first randomisation period were not reported; one reported clinical data for the maintenance phase only, with no data for the acute phase; one study compared fluvoxamine against imipramine and a placebo, but clinical data were shown for comparison between antidepressants and the placebo, not between fluvoxamine and imipramine.

Risk of bias in included studies

See: Included studies, Figure 2, Figure 3.

Figure 2.

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

Figure 3.

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

Our judgment about the overall risk of bias in the individual studies is illustrated in Figure 2 and Figure 3. The methodological quality of these included studies was judged as poor, although judging articles from some time ago by today's standard (Begg 1996; CONSORT) is problematic. Nevertheless, the reporting in these studies was not good. This type of reporting has been associated with an overestimate of the estimate of effect (Schulz 1995) and this should be considered when interpreting the results.

Allocation

Only one study reported the methods of generating random sequence, in which "a computer originated schedule" was used (Rossini 2005). No studies reported the method of allocation concealment. We were not assured that bias was minimised during the allocation procedure, yet 28 studies reported that the participants allocated to each treatment group were "similar", "the same", "not significantly different", "comparable" or "matched" (27/53 studies, 51%).

Blinding

Forty-three of the trials (80%) described their studies as "double blind"; however, no tests were conducted to ensure successful blinding. We rated only two studies among those 43 trials as having a "low risk of bias" (Miller 2001; Rossini 2005), as blinded raters conducted outcome assessment of those trials. In addition, one "single-blind" trial (Otsubo 2005) was rated as having a "low risk of bias" because its outcome assessment was blinded to the medication. Four trials were open trials that did not seek blinding (Barge-Schaapveld 1995; Brunner 1994; Kato 2006; Mendonca Lima 1997).

Incomplete outcome data

Total dropout rate was relatively high, ranging from 0% (de Wilde 1983; Mendonca Lima 1997; Miller 2001; Tourigny-Rivard 1996) to 59 % (Claghorn 1996). There were twenty-seven studies (27/54, 50%) where the total dropout rates were more than 20%.

Selective reporting

The study protocol was not available for all studies. Four studies reported only "pituitary-adrenocortical status" (Brown 1986), "prolactin response to d-fenfluramine challenge" (Kavoussi 1999), "heart rate" (Rechlin 1994), or "hypothalamic-pituitary-adrenocortical axis activity" (Rota 2005) instead of reporting the clinical outcome for each intervention group. One study reported the clinical efficacy outcome only as "we could not find any significant difference" (Gonul 1999). Only ten studies reported SDs of change scores (Asakura 2005; Harris 1991a; Kato 2006; Kiev 1997; Mendonca Lima 1997; Miller 2001; Nemeroff 1995; Otsubo 2005; Schoemaker 2002; Ueda 2002); 26 studies reported SDs of endpoint score of continuous efficacy variables.

Other potential sources of bias

Funding and wish bias

Most of the included studies (38 studies) were funded by industry. Among the 30 trials comparing fluvoxamine to TCAs, a great majority (21 trials) were sponsored by, or had at least one author affiliated with, a pharmaceutical company marketing fluvoxamine, and almost all of the trials (25 trials) set fluvoxamine as an investigational drug. Among the 24 trials comparing fluvoxamine with ADs other than TCAs, pharmaceutical companies marketing fluvoxamine sponsored 8 trials, and a company marketing the comparator drug funded 9 trials; only three trials set fluvoxamine as an investigational drug.

Effects of interventions

Of the 54 included studies (56 comparisons), 48 RCTs (50 comparisons) contributed usable data for the efficacy analyses and 49 RCTs (53 comparisons) did so for the tolerability analyses. No studies reported social adjustment/ functioning, health-related quality of life, and costs to health care services. ORs for the efficacy data larger than one (falling to the right of the midline) and those for the tolerability data smaller than one indicate a difference in favour of fluvoxamine. Negative SMDs (falling to the left of the midline) indicate a difference in favour of fluvoxamine.

To obtain response rate and remission, we used validated imputation methods, and if SDs were missing, we borrowed from other trials, if possible (See; Table 1, Table 2)

Table 1. Imputation methods and borrowed SD use to obtain response rate
 Early phaseAcute phase
Comparator StudyImputation methodsborrowed SDImputation methodsborrowed SD
TCAsImipramineCassano 1986YesYes--
Amore 1989YesYes--
Lydiard 1989YesYesYesYes
Bramanti 1988YesNo--
Claghorn 1996YesYesYesYes
Fabre 1992YesYesNoYes
Feighner 1989YesYesYesYes
Guy 1984YesYesYesYes
Itil 1983YesNo--
Koetsier 2002YesNo--
March 1990YesYesYesYes
Miller 2001----
ClomipramineColeman 1982YesYes--
de Wilde 1983YesYesYesYes
Dick 1983NoNo--
Ottevanger 1995YesNo--
Zohar 2003YesYesNoYes
AmitriptylineB-Schaapveld 1995NoNoNoNo
Harris 1991YesYesYesYes
Remick 1994YesNoYesNo
Kostiukova 2003NoYesNoYes
Murasaki 1998NoNo--
NortriptylineOtsubo 2005YesNoNoNo
DothiepinMullin 1988YesYesYesYes
Rahman 1991YesYesYesYes
DesipramineNathan 1990YesNo--
Tourigny Rivard 1996--YesYes
HeterocyclicsAmineptineBrunner 1994YesNo--
MianserinMoon 1991NoNoNoNo
Perez 1990YesYesYesYes
MaprotolineKasper 1989YesNo--
Mendonca Lima 1997YesYes--
SSRIsParoxetineAnnseaau 1993NoNoNoNo
Kato 2006NoYesNoYes
Kiev 1997YesYesYesNo
SertralineNemeroff 1995YesYesYesNo
Rossini 2002YesNoYesNo
FluoxetineDalery 1998NoYesNoYes
Rapaport 1995YesYesYesYes
CitalopramHaffmans 1996YesYesNoYes
SNRIMilnacipranClerc 2001NoNoNoNo
Ansseau 1991bNoNo--
Ansseau 1991aNoNo--
VenlafaxineHackett 1998aYesNoYesNo
Hackett 1998bYesNoYesNo
Newer ADsMirtazapineSchoemaker 2002NoYesNoYes
MoclobemideBarrelet 1991YesNo--
Bocksberger 1992YesNo--
Bougerol 1992YesNo--
Other conventionalSulpirideUeda 2002YesNo--
Table 2. Imputation methods and borrowed SD use to obtain remission rate
 Early phaseAcute phase
Comparator StudyImputation methodsborrowed SDImputation methodsborrowed SD
TCAsImipramineCassano 1986YesYes--
Amore 1989YesYes--
Lydiard 1989YesYesYesYes
Bramanti 1988YesNo--
Claghorn 1996YesYesYesYes
Fabre 1992YesYesYesYes
Feighner 1989YesYesYesYes
Guy 1984YesYesYesYes
Itil 1983NoNo--
Koetsier 2002YesNo--
March 1990YesYesYesYes
Miller 2001----
ClomipramineColeman 1982YesYes--
de Wilde 1983YesYesYesYes
Dick 1983NoNo--
Ottevanger 1995YesNo--
Zohar 2003YesYesYesYes
AmitriptylineB-Schaapveld 1995NoNoNoNo
Harris 1991YesYesYesYes
Remick 1994YesNoYesNo
Kostiukova 2003YesYesYesYes
Murasaki 1998YesNo--
NortriptylineOtsubo 2005YesNoNoNo
DothiepinMullin 1988YesYesYesYes
Rahman 1991YesYesYesYes
DesipramineNathan 1990YesNo--
Tourigny Rivard 1996--YesYes
HeterocyclicsAmineptineBrunner 1994YesNo--
MianserinMoon 1991NoNoNoNo
Perez 1990YesYesYesYes
MaprotolineKasper 1989YesNo--
Mendonca Lima 1997YesYes--
SSRIsParoxetineAnnseaau 1993YesNoYesNo
Kato 2006YesYesYesYes
Kiev 1997YesYesYesNo
SertralineNemeroff 1995YesYesYesNo
Rossini 2002NoNoNoNo
FluoxetineDalery 1998YesYesYesYes
Rapaport 1995YesYesYesYes
CitalopramHaffmans 1996YesYesNoYes
SNRIMilnacipranClerc 2001YesNoYesNo
Ansseau 1991bYesNo--
Ansseau 1991aYesNo--
VenlafaxineHackett 1998aNoNoNoNo
Hackett 1998bNoNoNoNo
Newer ADsMirtazapineSchoemaker 2002YesYesYesYes
MoclobemideBarrelet 1991YesNo--
Bocksberger 1992YesNo--
Bougerol 1992YesNo--
Other conventionalSulpirideUeda 2002YesNo--

1. FLUVOXAMINE versus TCAs

Twenty-eight RCTs contributed usable data for the efficacy analyses and 28 RCTs for the tolerability analyses. Twenty-one trials reported dichotomous data for a number of patients who experienced each side effect.

1.1 Response - acute phase (between 6 and 12 weeks); Primary outcome

Sixteen studies reported this outcome. Among them, imputation methods were used for 11 studies (Claghorn 1996; de Wilde 1983; Feighner 1989; Guy 1984; Harris 1991a; Lydiard 1989; March 1990; Mullin 1988; Rahman 1991; Remick 1994; Tourigny-Rivard 1996). There was no strong evidence that fluvoxamine was either superior or inferior to TCAs except desipramine in terms of this dichotomous outcome in head-to-head comparisons. However, desipramine were less effective than fluvoxamine based on one small trial (OR: 4.22, 95% CI 0.98 to 18.13, P=0.05; 1 trial, 47 participants) (Analysis 1.1, Figure 4).

Figure 4.

Forest plot of comparison: 1 Fluvoxamine vs TCAs, outcome: 1.1 Response (acute phase).

1.2 Response - early phase and follow-up phase
1.2.1 Early phase (between 1 and 4 weeks)

No strong evidence indicated that fluvoxamine was either superior or inferior to TCAs in terms of the dichotomous outcome between fluvoxamine and TCAs in head-to-head comparisons (Figure 5).

Figure 5.

Forest plot of comparison: 1 Fluvoxamine vs TCAs, outcome: 1.2 Response (early phase).

Substantial heterogeneity existed between trials comparing fluvoxamine to amitriptyline, based on four trials (I2= 70 %, P = 0.02, Analysis 1.2). Visual inspection revealed that, among these studies, three smaller ones using the imputation methods for response (Harris 1991a; Kostiukova 2003; Remick 1994) reported results favourable to amitriptyline. However, because of the small number of trials, sources of the heterogeneity cannot be further explained.

1.2.2 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

1.3 Remission
1.3.1 Early phase (between 1 and 4 weeks)

In terms of remission at the end of the early phase, the analysis found no strong evidence that fluvoxamine was either superior or inferior to TCA in head-to-head comparisons (Analysis 1.3).

1.3.2 Acute phase (between 6 and 12 weeks)

Fluvoxamine was found to be more effective than desipramine based on one trial (OR: 4.50, 95% CI 1.31 to 15.42, P=0.02; 1 trial, 47 participants) (Tourigny-Rivard 1996); this small study did not report the actual number of patients who experienced remission, so we converted HRSD data expressed as mean and SD into the number of remitted patients, according to the validated imputation method (Furukawa 2005) (Figure 6).

Figure 6.

Forest plot of comparison: 1 Fluvoxamine vs TCAs, outcome: 1.4 Remission (acute phase).

1.3.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

1.4 Endpoint score on depression scale
1.4.1 Early phase (between 1 and 4 weeks)

We meta-analysed non-skewed data only from 5 trials, and no strong evidence emerged that fluvoxamine was either superior or inferior to TCA as a class or in head-to-head comparisons (Analysis 1.5). However, data were skewed in six trials, and SDs were missing in 10 trials. We did not meta-analyse these data, and presented them descriptively (Analysis 1.6).

1.4.2 Acute phase (between 6 and 12 weeks)

Sixteen trials reported this outcome. However, we did not meta-analyse these data (data were skewed in three trials, and SDs were missing in 13 trials). We presented them descriptively (Analysis 1.7).

1.4.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

1.5. Change score on depression scale
1.5.1 Early phase (between 1 and 4 weeks)

We meta-analysed this outcome and found evidence that fluvoxamine was inferior to amitriptyline (SMD: 1.17, 95% CI 0.61 to 1.73, P<0.0001; 1 trial, 58 participants) (Analysis 1.8). However, four other trials that compared fluvoxamine with amitriptyline did not report SDs for this outcome and we did not meta-analyse these data (Analysis 1.9).

1.5.2 Acute phase (between 6 and 12 weeks).

We meta-analysed non-skewed data only from 3 trials, and there was no strong evidence that fluvoxamine was superior or inferior to TCA as a class or in head-to-head comparisons (Analysis 1.10). However, SDs were missing in 14 trials and we did not meta-analyse these data, and presented them descriptively instead (Analysis 1.11).

1.5.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

1.6 Tolerability
1.6.1 Dropout

No strong evidence emerged that fluvoxamine was more or less acceptable in terms of the total numbers of dropouts for any reason, a proxy measure of tolerability, between fluvoxamine and TCAs in head-to-head comparisons (Figure 7). Similarly, regarding patients who dropped out because of inefficacy, no strong evidence emerged that fluvoxamine was superior or inferior to TCAs in head-to-head comparisons (Analysis 1.13). The analysis of dropouts due to side effects revealed that amitriptyline (OR: 0.59, 95% CI 0.35 to 1.00, P=0.05; 5 trials, 420 participants) and TCA as a class (OR: 0.79, 95% CI 0.60 to 1.04, P=0.09; 21 trials, 1772 participants) were less tolerated than fluvoxamine (Analysis 1.14, Figure 8).

Figure 7.

Forest plot of comparison: 1 Fluvoxamine vs TCAs, outcome: 1.11 Total Dropout.

Figure 8.

Forest plot of comparison: 1 Fluvoxamine vs TCAs, outcome: 1.13 Dropout due to side effects.

1.6.2 Number of patients experiencing at least one side effect

People allocated to amitriptyline were more likely to have at least one side effect during the trial, even though it was not statistically significant (OR 0.66, 95%CI 0.42 to 1.04, P=0.07; 3 trials, 327 participants) (Analysis 1.15, Figure 9).

Figure 9.

Forest plot of comparison: 1 Fluvoxamine vs TCAs, outcome: 1.14 Number of patients experiencing at least one side effect.

1.7 Side effects profile by body system

See: Table 3.

Table 3. Side effect profiles by body system
Control drugBody systemSide effectN of comparisonsN of participantsOR95% CINNT95% CI
versus TCAs
ImipramineCardiovascularHypotension / bradycardia45600.240.10, 0.621613, 33
DermatologicalSweating79720.320.16, 0.661411, 28
GastrointestinalDry mouth910550.240.16, 0.3443, 5
Vomiting / nausea910552.231.59, 3.14-9-6, -17
Constipation810080.500.27, 0.93118, 86
Diarrhoea2136 6.381.27, 32.04-8-3, -133
NeuropsychiatricDizziness / vertigo / faintness910550.240.15, 0.3876, 8
Anxiety / agitation56442.241.01, 4.97-17-6, -1893
GenitourinaryProblems urinating24090.180.04, 0.711815, 51
ClomipramineGastrointestinalDry mouth32160.430.22, 0.8154, 20
Nausea / vomiting3216 2.131.06, 4.27-9-4, -138
NeuropsychiatricDizziness / vertigo /faintness1860.210.05, 0.8065, 25
AmitriptylineGastrointestinalVomiting / nausea4387 2.861.31, 6.23-13-6, -68
NeuropsychiatricDizziness / vertigo / faintness23040.310.11, 0.8375, 30
NortriptylineNeuropsychiatricDizziness / vertigo / faintness1740.220.07, 0.7043, 12
DothiepineGastrointestinalDry mouth1730.080.01, 0.7055, 17
versus Heterocyclics
MaprtilineNeuropsychiatricDizziness / vertigo / faintness1420.130.03, 0.5632, 6
MianserinGastrointestinalNausea/vomiting2125 9.621.96, 47.30-5-2, -36
versus SSRIs
ParoxetineDermatologicalSweating1600.220.05, 0.9154, 49
versus SNRIs
MilnacipranGastrointestinalVomiting / nausea3241 1.951.09, 3.50-7-4, -83
versus newer ADs
MirtazapineGastrointestinalVomiting / nausea1412 3.431.90, 6.19-7-4, -16
NeuropsychiatricSleepiness / drowsiness14120.470.29, 0.7686, 19
Agitation / anxiety14120.170.05, 0.611614, 35
MoclobemideGastrointestinalDry mouth2191 4.731.14, 19.57-15-4, -336
Vomiting / nausea2170 2.011.03, 3.92-7-4, -178
Only results for statistically significant difference were shown. OR, odds ratio. OR < 1 favours fluvoxamine.

All specific side effects were grouped by organ system, as follows:
Cardiovascular: hypertension/tachycardia, hypotension/bradycardia.
Dermatological: dermatitis/rash, sweating.
Gastrointestinal: increased salivation, dry mouth, oral discomfort/taste disturbance, vomiting/nausea, constipation, diarrhoea, weight gain, weight loss, increased appetite, anorexia.
Neruopsychiatric: blurred vision, dizziness/vertigo/faintness, fatigue/tiredness/asthenia, headache, tremor, involuntary movement other than tremor, insomnia, sleepiness/drowsiness, agitation/anxiety, manic symptom, completed suicide, suicide wishes/gestures/attempts.
Genitourinary: problems urinating, sexual dysfunction.

1.7.1 Cardiovascular side effects

Reasonable evidence indicated that fluvoxamine was less likely to cause hypotension / bradycardia than was imipramine (OR 0.24, 95%CI 0.10 to 0.62, P=0.003; 4 trials, 560 participants) (Analysis 7.2). No strong evidence suggested that fluvoxamine was more or less likely to cause hypertension/tachycardia than TCAs.

1.7.2 Dermatological side effects

Sweating was more frequent in imipramine-treated patients (OR 0.32, 95%CI 0.16 to 0.66, P=0.002; 7 trials, 972 participants) than in fluvoxamine-treated patients (Analysis 7.4). No strong evidence was apparent to indicate that fluvoxamine was more or less likely to cause dermatitis/rash than were TCAs.

1.7.3 Gastrointestinal side effects

Dry mouth was more frequent in patients treated with imipramine (OR 0.24, 95%CI 0.16 to 0.34, P<0.001; 9 trials, 1055 participants), clomipramine (OR 0.43, 95%CI 0.22 to 0.81, P=0.009; 7 trials, 972 participants), and dothiepin (OR 0.08, 95%CI 0.01 to 0.70, P=0.02; 1 trial, 972 participants) than in those treated with fluvoxamine (Analysis 7.6). Constipation was more frequent in patients treated with imipramine (OR 0.50, 95%CI 0.27 to 0.93, P=0.03; 8 trials, 1008 participants) (Analysis 7.9).

In contrast, fluvoxamine was associated with higher rates of vomiting/nausea in participants than occurred with imipramine (OR 2.23, 95%CI 1.59 to 3.14, p<0.001; 9 trials, 1055 participants), clomipramine (OR 2.13, 95%CI 1.06 to 4.27, P=0.03; 3 trials, 216 participants), and amitriptyline (OR 2.86, 95%CI 1.31 to 6.23, P=0.008; 4 trials, 387 participants) (Analysis 7.8). Diarrhoea was more frequent in patients treated with fluvoxamine than in those treated with imipramine (OR 6.38, 95%CI 1.27 to 32.04, P=0.02; 2 trials, 136 participants) (Analysis 7.10).

In terms of the rate of other gastrointestinal side effects (i.e., increased salivation, oral discomfort/taste disturbance, weight gain, weight loss, increased appetite or anorexia), no strong evidence emerged that fluvoxamine was either more or less likely to cause these adverse events than were TCAs.

1.7.4 Neuropsychiatric side effects

Dizziness/vertigo/faintness were less common in recipients of fluvoxamine than in recipients of imipramine (OR 0.24, 95%CI 0.15 to 0.38, p<0.001; 9 trials, 1055 participants), clomipramine (OR 0.21, 95%CI 0.05 to 0.80, P=0.02; 1 trial, 86 participants), amitriptyline (OR 0.31, 95%CI 0.11 to 0.83, P=0.02; 2 trials, 304 participants), nortriptyline (OR 0.22, 95%CI 0.07 to 0.70, P=0.01; 1 trial, 73 participants) (Analysis 7.16).

In contrast, fluvoxamine was associated with higher rate of agitation/anxiety in participants than was imipramine (OR 2.24, 95%CI 1.01 to 4.97, P=0.05; 5 trials, 644 participants) (Analysis 7.23).

Recent research has pointed out that some antidepressants, in particular SSRIs, have caused the emergence or worsening of suicidal ideas in vulnerable patients (Barbui 2008; Hammad 2006). Only two trials (Dick 1983; Zohar 2003) among those comparing fluvoxamine with TCAs recorded completed suicide (Analysis 7.25), with two events among 61 patients taking fluvoxamine and no events among 57 those taking TCAs. Suicide wishes/gestures/attempts were reported in only three trials (Cassano 1986; Mullin 1988; Zohar 2003), with 3 events among 250 patients taking fluvoxamine and five among 239 patients taking TCAs.

In terms of the rate of participants experiencing other neuropsychiatric side effects (i.e., blurred vision, fatigue/tiredness/asthenia, headache, tremor, involuntary movement other than tremor, insomnia, sleepiness or manic symptoms), no strong evidence emerged that fluvoxamine was either more or less likely to cause these adverse events than were TCAs.

1.7.5 Genitourinary side effects

Urination problems were less common in recipients of fluvoxamine than in recipients of imipramine (OR 0.18, 95%CI 0.04 to 0.71, P=0.01; 2 trials, 409 participants) and TCAs as a class (OR 0.44, 95%CI 0.23 to 0.83, P=0.01; 6 trials, 818 participants) (Analysis 7.27). In terms of the rate of participants experiencing sexual dysfunction, no strong evidence emerged that fluvoxamine was more or less likely to cause these adverse events than were TCAs.

2. FLUVOXAMINE versus HETEROCYCLICS

Only five RCTs contributed usable data for efficacy and tolerability analyses (versus amineptine: Brunner 1994, versus maprotiline: Kasper 1990; Mendonca Lima 1997, versus mianserin: Moon 1991; Perez 1990). Four trials reported dichotomous data for a number of patients who experienced each side effect (versus maprotiline: Kasper 1990; Mendonca Lima 1997, versus mianserin: Moon 1991; Perez 1990).

2.1 Response - acute phase (between 6 and 12 weeks); Primary outcome

Two trials comparing fluvoxamine with mianserin reported this outcome. Imputation methods were used for Perez 1990. No strong evidence emerged that fluvoxamine was either superior or inferior to mianserin in terms of the response at end of the acute-phase treatment (Analysis 2.1).

2.2 Response - early phase and follow-up phase
2.2.1 Early phase (between 1 and 4 weeks)

No strong evidence emerged that fluvoxamine was either superior or inferior to amineptine or maprotiline in terms of the response at end of the acute-phase treatment. No trials comparing fluvoxamine with mianserin reported this outcome. See Analysis 2.2

2.2.2 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

2.3 Remission
2.3.1 Early phase (between 1 and 4 weeks)

No strong evidence emerged that fluvoxamine was either superior or inferior to amineptine or maprotiline in terms of the remission . No trials comparing fluvoxamine with mianserin reported this outcome. See Analysis 2.3

2.3.2 Acute phase (between 6 and 12 weeks)

Two trials that compared fluvoxamine with mianserin reported this outcome. No strong evidence emerged to indicate that fluvoxamine was either superior or inferior to mianserin in terms of this outcome. See Analysis 2.4.

2.3.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

2.4 Endpoint score on depression scale
2.4.1 Early phase (between 1 and 4 weeks)

We meta-analysed non-skewed data only from 2 trials, and found no strong evidence that fluvoxamine was either superior or inferior to heterocyclics (Analysis 2.5). SDs were missing in one trial and we did not meta-analyse these data, presenting them descriptively instead (Analysis 2.6).

2.4.2 Acute phase (between 6 and 12 weeks)

Only one trial (Perez 1990) reported this continuous outcome, but SDs were missing (Analysis 2.7).

2.4.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

2.5 Change score on depression scale
2.5.1 Early phase (between 1 and 4 weeks)

We meta-analysed non-skewed data only from 1 trial and found no strong evidence that fluvoxamine was either superior or inferior to moprotiline (Analysis 2.8). SDs were missing in three trials and we did not meta-analyse these data, presenting them descriptively instead (Analysis 2.9).

2.5.2 Acute phase (between 6 and 12 weeks)

Two trials compared fluvoxamine with mianserin and reported this continuous outcome. However, SDs for this outcome were missing. Therefore, we presented the results descriptively (Analysis 2.10).

2.5.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

2.6 Tolerability
2.6.1 Dropout

We found no strong evidence that fluvoxamine was either more or less acceptable in terms of the total numbers of dropouts for any reason when compared to heterocyclics (i.e., amineptine, maprotiline and mianserin) (Analysis 2.11). Similarly, regarding patients who dropped out because of inefficacy and due to side effects, we found no strong evidence that fluvoxamine was superior or inferior to heterocyclics (Analysis 2.12, Analysis 2.13).

2.6.2 Number of patients experiencing at least one side effect

No strong evidence emerged that fluvoxamine was either superior or inferior to maprotiline or mianserin in terms of this dichotomous outcome (Analysis 2.14).

2.7 Side effects profile by body system

No trials comparing fluvoxamine with amineptine reported a number of patients who experienced a specific side effect. Only gastrointestinal side effects, such as dry mouth, vomiting/nausea and dizziness/vertigo/faintness were reported by four RCTs (Kasper 1990; Mendonca Lima 1997; Moon 1991; Perez 1990). See: Table 3.

2.7.1 Cardiovascular side effects

No studies contributed data to this outcome.

2.7.2 Dermatological side effects

No studies contributed data to this outcome.

2.7.3 Gastrointestinal side effects

Reasonable evidence existed that fluvoxamine was associated with a higher rate of vomiting/nausea than was mianserin (OR 9.62, 95%CI 1.96 to 47.30, p=0.009; 4 trials, 207 participants). For further details, see Analysis 8.1, Analysis 8.2, Analysis 8.3). In terms of dry mouth or dizziness/vertigo/faintness, we found no strong evidence that fluvoxamine was either more or less likely to cause these adverse events than were heterocyclics.

2.7.4 Neuropsychiatric side effects

No studies contributed data to this outcome.

2.7.5 Genitourinary side effects

No studies contributed data to this outcome.

3. FLUVOXAMINE versus OTHER SSRIs

Eight RCTs contributed usable data for the efficacy analyses and nine RCTs contributed to the tolerability analyses. Eight trials reported dichotomous data for a number of patients who experienced each side effect.

3.1 Response - acute phase between 6 and 12 weeks; Primary outcome

Eight trials reported this outcome. Imputation methods were used for four trials (Kiev 1997, Nemeroff 1995, Rossini 2005, Rapaport 1996). We found no strong evidence that fluvoxamine was either superior or inferior to other SSRIs (i.e., paroxetine, sertraline, fluoxetine and citalopram) in terms of this dichotomous outcome. See Figure 10.

Figure 10.

Forest plot of comparison: 3 Fluvoxamine vs other SSRIs, outcome: 3.1 Response (acute phase).

3.2 Response - early phase and follow-up phase
3.2.1 Early phase (between 1 and 4 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to other SSRIs in terms of this dichotomous outcome. See Analysis 3.2. Substantial heterogeneity existed between trials comparing fluvoxamine to fluoxetine based on two trials, Dalery 2003 and Rapaport 1996 ( I2= 71 %, P = 0.07, Analysis 3.2). However, because of the small number of trials, sources of the heterogeneity cannot be further explained.

3.2.2 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

3.3 Remission
3.3.1 Early phase (between 1 and 4 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to other SSRIs in terms of remission at the end of early phase. See Analysis 3.3.

3.3.2 Acute phase (between 6 and 12 weeks)

No strong evidence emerged that fluvoxamine was either superior or inferior to other SSRIs in terms of remission at the end of acute phase. See Analysis 3.4.

3.3.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

3.4 Endpoint score on depression scale
3.4.1 Early phase (between 1 and 4 weeks)

We meta-analysed non-skewed data only from 2 trials, and found no strong evidence that fluvoxamine was either superior or inferior to other SSRIs (Analysis 3.5). SDs were missing in five trials and we did not meta-analyse these data, and presenting them descriptively instead (Analysis 3.6).

3.4.2 Acute phase (between 6 and 12 weeks)

Eight trials reported this outcome. However, we did not meta-analyse these data (data were skewed in four trials, and SDs were missing in four trials). We presented them descriptively (Analysis 3.7).

3.4.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

3.5 Change score on depression scale
3.5.1 Early phase (between 1 and 4 weeks)

We meta-analysed data only from one trial, and found no strong evidence that fluvoxamine was either superior or inferior to paroxetine (Analysis 3.8). SDs were missing in six trials and we did not meta-analyse these data, and presenting them descriptively instead (Analysis 3.9).

3.5.2 Acute phase (between 6 and 12 weeks)

We meta-analysed data only from three trials, and found no strong evidence that fluvoxamine was either superior or inferior to paroxetine or sertraline (Analysis 3.10). SDs were missing in five trials and we did not meta-analyse these data, presenting them descriptively instead (Analysis 3.11).

3.5.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

3.6. Tolerability
3.6.1 Dropout

We found no strong evidence that fluvoxamine was either more or less acceptable than were other SSRIs in terms of withdrawal due to any reason (Analysis 3.12). Similarly, regarding number of patients who dropped out because of inefficacy and because of adverse effects, no strong evidence emerged that fluvoxamine was either superior or inferior to other SSRIs (Analysis 3.13 and Analysis 3.14). We found substantial heterogeneity between trials comparing fluvoxamine to sertraline, based on three trials (Gonul 1999; Nemeroff 1995; Rossini 2005) (I2 = 66 %, P = 0.05, Analysis 3.14.2). Visual inspection revealed that, among these studies, Nemeroff 1995 reported results favourable to sertraline. However, because of the small number of trials, sources of the heterogeneity cannot be further explained.

3.6.2 Number of patients experiencing at least one side effect

Only five trials reported this dichotomous outcome, and no strong evidence emerged that fluvoxamine was either superior or inferior to other SSRIs in terms of this outcome. See Analysis 3.15.

3.7 Side effect profile by body system

Only one trial (Haffmans 1996) compared fluvoxamine with citalopram but did not report the number of patients who experienced specific side effects other than completed suicide or suicide attempts. See: Table 3.

3.7.1 Cardiovascular side effects

In terms of the rate of participants experiencing hypertension/tachycardia or hypotension/bradycardia, we found no strong evidence that fluvoxamine was either superior or inferior to paroxetine (Analysis 9.1, Analysis 9.2).

3.7.2 Dermatological side effects

Sweating was more frequent with paroxetine (OR 0.22, 95%CI 0.05 to 0.91, P=0.04; 1 trial, 60 participants) than in fluvoxamine-treated patients (Analysis 9.4). In terms of the rate of dermatitis/rash experienced by participants, no strong evidence emerged that fluvoxamine was either superior or inferior to sertraline in this respect (Analysis 9.3).

3.7.3 Gastrointestinal side effects

We found no evidence that fluvoxamine was either superior or inferior to other SSRIs in terms of the rate of gastrointestinal side effects experienced by participants (Analysis 9.5, Analysis 9.6, Analysis 9.7, Analysis 9.8 and Analysis 9.9).

No studies contributed data regarding increased salivation, oral discomfort/taste disturbance, weight gain, weight loss or increased appetite.

3.7.4 Neuropsychiatric side effects

We found no strong evidence that fluvoxamine, compared with other SSRIS, was either more or less likely to cause headache (Analysis 9.12), insomnia (Analysis 9.14), agitation/anxiety (Analysis 9.16), dizziness/vertigo/faintness (Analysis 9.10), fatigue/tiredness/asthenia (Analysis 9.11), tremor (Analysis 9.13), sleepiness/drowsiness (Analysis 9.15), manic symptom (Analysis 9.17).
Only one trial compared fluvoxamine with citalopram (Haffmans 1996) recorded a completed suicide (Analysis 7.25), and cited one event among 108 patients taking citalopram and no events among 109 patients taking fluvoxamine. Suicide attempts/ideation were reported in only four trials (Ansseau 1994; Dalery 2003; Rapaport 1996; Haffmans 1996), with 6 events among 314 patients taking fluvoxamine and 2 patients among 307 patients taking other SSRIs (i.e., paroxetine, fluoxetine and citalopram) (Analysis 9.18, Analysis 9.19).
No studies contributed data to blurred vision and involuntary movement other than tremor.

3.7.5 Genitourinary side effects

Only four trials reported sexual dysfunction as a side effect. Some previous trials have reported that fluvoxamine was associated with a relatively low prevalence of sexual dysfunction compared to other SSRIs (i.e., paroxetine, sertraline and fluoxetine) ( (Mackay 1997; Montejo-Gonzalez 1997). However, we found no strong evidence that fluvoxamine was either more or less likely to cause sexual dysfunction than were other SSRIs (Analysis 9.20). No studies contributed data to problems urinating.

4. FLUVOXAMINE versus SNRIs

Three RCTs (five comparisons) contributed usable data for the efficacy analyses and tolerability analyses for milnacipran or venlafaxine (Ansseau 1991a; Ansseau 1991b; Clerc 2001; Hackett 1998a; Hackett 1998b). Two RCTs (three comparisons) reported dichotomous data for a number of patients who experienced specific side effects (Ansseau 1991a; Ansseau 1991b; Clerc 2001).

4.1 Response - acute phase (between 6 and 12 weeks); Primary outcome

Two trials (three comparisons) reported this outcome. Imputation methods were used for Hackett 1998a and Hackett 1998b. Evidence emerged that fluvoxamine was less effective than was venlafaxine, based on one trial (2 comparisons), for which the response had to be calculated based on the imputation method (Furukawa 2005) (OR: 0.40, 95% CI 0.18 to 0.92, P=0.03; 1 trial (2 comparisons), 111 participants) (Hackett 1998a; Hackett 1998b). We found no strong evidence that fluvoxamine was either superior or inferior to milnacipran in terms of this dichotomous outcome. See Figure 11.

Figure 11.

Forest plot of comparison: 4 Fluvoxamine vs SNRIs, outcome: 4.1 Response (acute phase).

4.2 Response - early phase and follow-up phase
4.2.1 Early phase (between 1 and 4 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to SNRIs in terms of this dichotomous outcome. See Analysis 4.2.

4.2.2 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

4.3 Remission
4.3.1 Early phase (between 1 and 4 weeks)

No strong evidence emerged that fluvoxamine was either superior or inferior to SNRIs in terms of this dichotomous outcome. See Analysis 4.3.

4.3.2 Acute phase (between 6 and 12 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to SNRIs in terms of this dichotomous outcome. See Analysis 4.4.

4.3.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

4.4 Endpoint score on depression scale
4.4.1 Early phase (between 1 and 4 weeks)

We meta-analysed data only from 3 trials (four comparisons), and found no strong evidence that fluvoxamine was either superior or inferior to mianserin or venlafaxine (Analysis 4.5). SDs were missing in one trial and we did not meta-analyse the data, and presenting it descriptively instead (Analysis 4.6).

4.4.2 Acute phase (between 6 and 12 weeks)

One trial (two comparisons) reported this outcome. However, we did not meta-analyse these skewed data, instead presenting them descriptively (Analysis 4.7).

4.4.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

4.5 Change score on depression scale
4.5.1 Early phase (between 1 and 4 weeks)

Three trials (five comparisons) reported this outcome. However, we did not meta-analyse these data, since SDs were missing (Analysis 4.8).

4.5.2 Acute phase (between 6 and 12 weeks)

Two trials (three comparisons) reported this outcome. However, we did not meta-analyse these data, since SDs were missing (Analysis 4.9).

4.5.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

4.6 Tolerability
4.6.1 Dropout

We found no strong evidence that fluvoxamine was either more or less acceptable in terms of withdrawal due to any reason when compared with milnacipran. Venlafaxine was found to be less likely elicit withdrawal due to any reason (OR: 2.29, 95% CI 0.97 to 5.43, P=0.06; 1 trial (2 comparisons), 111 participants) (Analysis 4.10). Similarly, regarding the number of patients who dropped out because of inefficacy and due to side effects, we found no strong evidence that fluvoxamine was either superior or inferior to milnacipran. (Analysis 4.11, Analysis 4.12).

4.6.2 Number of patients experiencing at least one side effect

No studies contributed data to this outcome.

4.7 Side effects profile by body system

No studies comparing fluvoxamine to venlafaxine reported any number of patients experiencing specific side effects. See: Table 3.

4.7.1 Cardiovascular side effects

In terms of the rate of hypertension/tachycardia or hypotension/bradycardia, we found no strong evidence that fluvoxamine was either superior or inferior to milnacipran in this respect (Analysis 10.1, Analysis 10.2).

4.7.2 Dermatological side effects

In terms of the rate of dermatitis/rash or sweating experienced by participants, no strong evidence emerged that fluvoxamine was either superior or inferior to milnacipran. (Analysis 10.3, Analysis 10.4).

4.7.3 Gastrointestinal side effects

We found evidence that fluvoxamine was associated with higher rate of vomiting/nausea experienced by participants than was milnacipran (OR 1.95, 95%CI 1.09 to 3.50, P=0.02; 2 trials (3 comparisons), 240 participants) (Analysis 10.8). In terms of the rate of increased salivation, dry mouth, oral discomfort/taste disturbance, vomiting/nausea, constipation, diarrhoea, weight gain, weight loss and anorexia experienced by participants receiving fluvoxamine compared to milnacipran, no strong evidence emerged to indicate that fluvoxamine was either superior or inferior to milnacipran. See Analysis 10.5 to Analysis 10.13.

No studies contributed data to increased appetite.

4.7.4 Neuropsychiatric side effects

In terms of the rate of blurred vision, dizziness/vertigo/faintness, fatigue/tiredness/asthenia, headache, tremor, involuntary movement other than tremor, insomnia, sleepiness/drowsiness and agitation/anxiety experienced by participants, we found no strong evidence that fluvoxamine was either superior or inferior to milnacipran. See Analysis 10.14 to Analysis 10.22.

No studies contributed data to manic symptom, completed suicide and suicide wishes/gestures/attempts.

4.7.5 Genitourinary side effects

We found no strong evidence that fluvoxamine was either more or less likely to cause urination problems than was milnacipran. See Analysis 10.23

No studies contributed data to sexual dysfunction.

5. FLUVOXAMINE versus NEWER ANTIDEPRESSANTS

Three RCTs comparing fluvoxamine to moclobemide (Barrelet 1991; Bocksberger 1993; Bougerol 1992) and one RCT comparing fluvoxamine to mirtazapine (Schoemaker 2002) contributed usable data for the efficacy analyses, tolerability analyses and for a number of patients who experienced specific side effects. These two drugs have little chemically in common, we did not pool the results of these two drugs, presenting them separately instead.

5.1 Response - acute phase (between 6 and 12 weeks); Primary outcome

Schoemaker 2002 reported this outcome. Imputation methods were not used for this study. No strong evidence emerged that fluvoxamine was either superior or inferior to mirtazapine in terms of this dichotomous outcome. See Analysis 5.1.

5.2 Response - early phase and follow-up phase
5.2.1 Early phase (between 1 and 4 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to moclobemide or mirtazapine in terms of this dichotomous outcome. See Analysis 5.2.

5.2.2 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

5.3 Remission
5.3.1 Early phase (between 1 and 4 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to moclobemide or mirtazapine in terms of this dichotomous outcome. See Analysis 5.3.

5.3.2 Acute phase (between 6 and 12 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to mirtazapine in terms of this dichotomous outcome. See Analysis 5.4.

5.3.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

5.4 Endpoint score on depression scale
5.4.1 Early phase (between 1 and 4 weeks)

Two trials reported this outcome. However, we did not meta-analyse these data, since SDs were missing (Analysis 5.5).

5.4.2 Acute phase (between 6 and 12 weeks)

Two trials reported this outcome. However, we did not meta-analyse these data, since SDs were missing (Analysis 5.6).

5.4.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

5.5 Change score on depression scale
5.5.1 Early phase (between 1 and 4 weeks)

One trial that compared fluvoxamine with mirtazapine reported this outcome and showed strong evidence that fluvoxamine was inferior to mirtazapine (SMD 0.32, 95%CI 0.12 to 0.51, P=0.002; 1 trial, n=402) (Analysis 5.7). Three trials that compared fluvoxamine with moclobemide reported this outcome, but we did not meta-analyse these data since SDs were missing (Analysis 5.8)

5.5.2 Acute phase (between 6 and 12 weeks)

One trial that compared fluvoxamine with mirtazapine reported this outcome but showed no strong evidence that fluvoxamine was either inferior or superior to mirtazapine (SMD 0.08, 95%CI 0.12 to 0.28, P=0.42; 1 trial, n=402) (Analysis 5.9).

5.5.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

5.6 Tolerability
5.6.1 Dropout

We found no strong evidence that fluvoxamine was either more or less acceptable in terms of withdrawal due to any reason when compared with the newer ADs (Analysis 5.10). Similarly, regarding number of patients who dropped out because of inefficacy and due to side effects, we found no strong evidence that fluvoxamine was either more or less acceptable than the newer ADs (Analysis 5.11, Analysis 5.12).

5.6.2 Number of patients experiencing at least one side effect

Fluvoxamine did appear to be associated with a higher number of participants experiencing at least one side effect when compared with moclobemide (OR 2.29, 95%CI 1.35 to 3.88, P=0.002; 3 trials, 231 participants) (Analysis 5.13).

5.7 Side effects profile by body system

See: Table 3.

5.7.1 Cardiovascular side effects

In terms of the numbers of participants experiencing hypertension/tachycardia or hypotension/bradycardia, no strong evidence emerged that fluvoxamine was either more or less likely to cause these adverse events than was moclobemide (Analysis 11.1). No studies contributed data to hypertension/tachycardia.

5.7.2 Dermatological side effects

We found no strong evidence that fluvoxamine was either more or less likely to cause sweating than was moclobemide (Analysis 11.2). No studies contributed data to dermatitis/rash.

5.7.3 Gastrointestinal side effects

Fluvoxamine appeared to be associated with a higher number of participants who experienced vomiting/nausea when compared with mirtazapine (OR 3.43, 95%CI 1.90 to 6.19, P<0.001; 1 trial, 412 participants) or moclobemide (OR 2.01, 95%CI 1.03 to 3.92, P=0.04; 2 trials, 170 participants) (Analysis 11.4). We found evidence that fluvoxamine was associated with a higher number of participants who experienced dry mouth when compared with mirtazapine (OR 4.73, 95%CI 1.14 to 19.57, P=0.03; 1 trial, 412 participants) (Analysis 11.3).

No strong evidence emerged that fluvoxamine, compared with mirtazapine, was either more or less likely to cause constipation, diarrhoea, weight gain or increased appetite. See Analysis 11.5 to Analysis 11.8.

No studies contributed data to increased salivation, oral discomfort/taste disturbance, weight loss or anorexia.

5.7.4 Neuropsychiatric side effects

See Analysis 11.9 to Analysis 11.18.

Fluvoxamine appeared to be associated with lower numbers of participants who experienced sleepiness/drowsiness when compared with mirtazapine (OR 0.47, 95%CI 0.29 to 0.76, P=0.002; 1 trial, 412 participants) (Analysis 11.15). Fluvoxamine was associated with lower numbers of participants who experienced agitation/ anxiety when compared with mirtazapine (OR 0.17, 95%CI 0.05 to 0.61, P=0.03; 1 trial, 412 participants) (Analysis 11.16).
Only one trial compared fluvoxamine with moclobemide (Barrelet 1991) and recorded completed suicide (Analysis 11.18), with one event among 30 patients taking fluvoxamine and no events among 31 patients taking moclobemide.

No studies contributed data to involuntary movements other than tremors or to suicide wishes/gestures/attempts.

5.7.5 Genitourinary side effects

No studies contributed data to genitourinary side effects.

6. FLUVOXAMINE versus OTHER CONVENTIONAL PSYCHOTROPIC DRUGS

Only one 4-week RCT comparing fluvoxamine to sulpiride (Ueda 2002) contributed usable data for the efficacy analyses and tolerability analyses.

6.1 Response - acute phase (between 6 and 12 weeks); Primary outcome

No studies contributed data to this outcome.

6.2 Response - early phase and follow-up phase
6.2.1 Early phase (between 1 and 4 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to sulpiride in terms of response at end of the acute-phase treatment. (Analysis 6.1).

6.2.2 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

6.3 Remission
6.3.1 Early phase (between 1 and 4 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to sulpiride in terms of remission at end of the early phase treatment. (Analysis 6.2).

6.3.2 Acute phase (between 6 and 12 weeks)

No studies contributed data to this outcome.

6.2.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

6.4 Endpoint score on depression scale
6.4.1 Early phase (between 1 and 4 weeks)

We found no strong evidence that fluvoxamine was either superior or inferior to sulpiride in terms of this continuous outcome (Analysis 6.3).

6.4.2 Acute phase (between 6 and 12 weeks)

No studies contributed data to this outcome.

6.4.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

6.5 Change score on depression scale
6.5.1 Early phase (between 1 and 4 weeks)

We found no strong evidence that fluvoxamine was either inferior or superior to sulpiride in terms of this continuous outcome (Analysis 6.4).

6.5.2 Acute phase (between 6 and 12 weeks)

No studies contributed data to this outcome.

6.5.3 Follow-up phase (between 4 and 6 months)

No studies contributed data to this outcome.

6.7 Tolerability
6.7.1 Dropout

We found no strong evidence that fluvoxamine was either more or less acceptable than was sulpiride in terms of withdrawal due to any reason (Analysis 6.5). No studies contributed data to dropout due to inefficacy or due to side effects.

6.7.2 Numbers of patients experiencing at least one side effect

No studies contributed data to this outcome.

6.8 Side effects profile by body system

Ueda 2002 did not report specific side effects.

7. FUNNEL PLOT ANALYSES

Funnel plots were examined only for the comparison between fluvoxamine and TCAs as a class, since there were insufficient trials to allow meaningful formal assessment using funnel plots for other comparisons. Visual inspection did not reveal an asymmetrical appearance of the funnel plot (Figure 12). Tests for funnel plot asymmetry (Egger 1997) did not suggest any strong evidence of asymmetry (Egger's bias coefficient, number of trials= 16, bias = 0.09 (P = 0.94)). However, included trials were of similar size (sample size: 23 to 100) and similar standard errors of OR (SE of ln(OR): 0.40 to 0.89), and the test for funnel plots should not be used (Higgins 2008).

Figure 12.

Funnel plot of comparison: 1 Fluvoxamine versus TCAs, outcome: 1.1 Response (acute phase): Primary outcome.

8. SUBGROUP ANALYSES

We conducted subgroup analyses only for fluvoxamine dosing - standard versus high dose (fluvoxamine versus TCAs, fluvoxamine versus other SSRIS), comparator dosing (fluvoxamine versus TCAs, fluvoxamine versus other SSRIS) and treatment settings (fluvoxamine versus TCAs), since there were insufficient number of trials that could produce subgroups and useful findings. Differences in subgroups were reported only for response rates in acute phases.

8.1 Fluvoxamine dosing - standard dose versus high dose
8.1.1 Fluvoxamine versus TCAs

Only 6 trials used standard dose schedules, and 10 trials used high dose schedules. The magnitude of effects in the trials using standard doses of fluvoxamine (OR: 0.83, 95% CI 0.53 to 1.46) and high doses of fluvoxamine (OR: 1.07, 95% CI 0.74 to 1.55) were similar and their CIs were overlapped (Analysis 1.16, Analysis 1.17).

8.1.2 Fluvoxamine versus other SSRIs

Only 5 trials used standard dose schedules, and 3 trials used high dose schedules. The magnitude of effects in the trials using standard doses of fluvoxamine (OR: 0.92, 95% CI 0.65 to 1.30) and high doses of fluvoxamine (OR: 1.02, 95% CI 0.65 to 1.60) were similar and their CIs were overlapped (Analysis 3.16, Analysis 3.17).

8.2 Comparator dosing
8.2.1 Fluvoxamine versus TCAs

Only 4 trials used standard dose schedules and 15 trials used high dose schedules of comparator drugs. The magnitude of effects in the trials using standard dose comparator drugs (OR: 1.27, 95% CI 0.64 to 2.50) and high dose comparator drugs (OR: 0.99, 95% CI 0.74 to 1.33) were similar and their CIs were overlapped (Analysis 1.18, Analysis 1.19).

8.2.2 Fluvoxamine versus other SSRIs

Only 2 trials used standard dose schedules, and 6 trials used high dose schedules of comparator drugs. The magnitude of effects in the trials using standard dose comparator drugs (OR: 0.95, 95% CI 0.60 to 1.50) and high dose of comparator drugs (OR: 0.96, 95% CI 0.70 to 1.33) were similar and their CIs were overlapped (Analysis 3.18, Analysis 3.19).

8.3 Treatment settings
8.3.1 Fluvoxamine versus TCAs

Only 5 trials were conducted in inpatient settings and 9 trials were conducted in outpatient settings. The magnitude of effects in the trials that were conducted in inpatient settings (OR: 1.17, 95% CI 0.71 to 1.92) and the trials conducted in outpatient settings (OR: 0.84, 95% CI 0.58 to 1.20) were similar and their CIs were overlapped (Analysis 1.20, Analysis 1.21).

9. SENSITIVITY ANALYSES

We reported the results of sensitivity analyses for efficacy outcome at the acute phase. In the main analyses, we found no evidence that fluvoxamine was either more or less effective than other ADs as a class. In head-to-head comparisons, fluvoxamine was more effective than desipramine (response rate at acute phase; OR: 4.22, 95% CI 0.98 to 18.13, P=0.05; 1 trial, 47 participants), while venlafaxine was more effective than fluvoxamine (response rate at acute phase; OR: 0.40, 95% CI 0.18 to 0.92, P=0.03; 1 trial (2 comparisons), 111 participants).

9.1 Excluding trials with unclear concealment of random allocation and/or unclear double blinding

We did not perform sensitivity analyses of excluded trials with unclear concealment of random allocation, as there were no studies that reported details of having conducted allocation concealment. In addition, we did not perform sensitivity analysis of excluding with unclear double blinding as there were only two studies rated as being "low risk of bias" regarding blinding.

9.2 Excluding trials with dropout rates greater than 20%

No substantial change was found in the main findings when fluvoxamine and TCAs (Analysis 1.22) or fluvoxamine and other SSRIs (Analysis 3.20) were compared for sensitivity analysis. We did not perform sensitivity analysis for comparisons between fluvoxamine and heterocyclics, SNRIs or newer ADs because dropout rates greater than 20 % were reported for all of the trials of these comparisons that reported primary outcome.

9.3 Performing the worst case scenario ITT

Based on sensitivity analysis, fluvoxamine was found to be less effective in terms of response at acute phase than were imipramine (OR 0.60 95%CI 0.39 to 0.90, P=0.01; 6 trials, 375 participants), amitriptyline (OR 0.34 95%CI 0.13 to 0.90, P=0.03; 4 trials, 185 participants), dothiepin (OR 0.34 95%CI 0.16 to 0.72, P=0.005; 2 trials, 125 participants) or venlafaxine (OR 0.11 95%CI 0.04 to 0.29, P<0.001; 1 trial (2 comparisons), 111 participants). See Analysis 1.23, Analysis 4.13.

9.4 Performing the best case scenario ITT

Based on sensitivity analysis, fluvoxamine was found to be more effective in terms of response at acute phase than were dothiepin (OR 4.04 95%CI 1.85 to 8.81, P<0.001; 2 trials, 125 participants) (Analysis 1.24), mianserin (OR 3.04 95%CI 1.20 to 7.67, P=0.02; 2 trials, 125 participants) (Analysis 2.16) or paroxetine (OR 1.77 95%CI 1.08 to 2.92, P=0.02; 3 trials, 281 participants) (Analysis 3.22). In addition, based on this sensitivity analysis, the superiority of venlafaxine over fluvoxamine in terms of response at acute phase was lost ( (OR 2.67 95%CI 0.98 to 7.26, P=0.05; 1 trial (2 comparisons), 111 participants) (Analysis 4.14).

9.5 Excluding trials for which the response rates had to be calculated based on the imputation method and for which the SD had to be borrowed from other trials

We conducted this sensitivity analysis only for comparison between fluvoxamine and TCAs, and between fluvoxamine and other SSRIs, since there were insufficient numbers of trials that could produce useful findings for other comparisons. No substantial change in the results was found by excluding trials with the imputation method for calculating responses (Analysis 1.25, Analysis 3.23) or by excluding trials that borrowed SDs for imputation (Analysis 1.26, Analysis 3.24).

9.6 Examination of "wish bias" by comparing fluvoxamine as an investigational drug versus fluvoxamine as a comparator

Examination of "wish bias" was impossible because no trials comparing fluvoxamine with TCAs set fluvoxamine as a comparator; among studies comparing fluvoxamine with ADs other than TCAs, only three trials set fluvoxamine as an investigational drug.

9.7 Excluding studies funded by or with at least one author affiliated with a pharmaceutical company marketing fluvoxamine

A sensitivity analysis to investigate the effect of commercial funding, excluding studies sponsored by pharmaceutical companies, was impossible, as almost all of the included trials had been funded by the industry. For example, among 30 trials comparing fluvoxamine with TCAs, there were only two trials free from commercial funding. Therefore, it is impossible to obtain substantial results from this sensitivity analysis.

9.8 Excluding studies that included patients with bipolar depression

No substantial changes in the main findings were noted by these sensitivity analyses (Analysis 1.30, Analysis 2.20, Analysis 3.28, Analysis 4.18).

9.9 Excluding studies that included patients with psychotic features

No substantial changes in the main findings were noted by these sensitivity analyses (Analysis 1.31, Analysis 2.21, Analysis 3.29, Analysis 4.19, Analysis 5.21).

Discussion

Summary of main results

The main results of our study revealed no strong evidence that fluvoxamine was either inferior or superior to other antidepressants, including TCAs such as amitriptyline or clomipramine, other SSRIs or other forms, in terms of either response or remission in any clinical setting, even though a small number of findings suggested a direction of effect in favour of fluvoxamine (versus desipramine) or in favour of control drug (venlafaxine) in terms of response and remission at acute phase. This was somewhat surprising because TCAs are sometimes believed to be more effective than SSRIs, particularly among hospitalized depressive patients (Anderson 1998).

We were also unable to find differences in dropouts for any reason, including side effects, between fluvoxamine and other antidepressive agents. Although the 95% CI are wide and we cannot exclude possibilities of meaningful differences between fluvoxamine and other antidepressive agents, our findings at least suggest a need for moderation of the general statement that holds that patients tolerate SSRIs better than TCAs. In addition, we included 10 trials involving fluvoxamine in our review, and from the pooled data, we were unable to find any differences in total dropouts or dropouts due to side effects between fluvoxamine and other SSRIs.

The analysis of individual side effects points to evidence of differing side effects profiles, especially when comparing gastrointestinal side effects between fluvoxamine and TCAs. Diarrhoea and weight loss were experienced significantly more frequently with fluvoxamine than with TCAs. Vomiting/nausea and weight loss/anorexia were also experienced significantly more frequently with fluvoxamine than with TCAs and some other antidepressive agents (e.g., mianserin, milnacipran and some newer antidepressants). However, constipation and decreased salivation/dry mouth were more common with TCAs than with fluvoxamine.

Not only are SSRIs chemically different from TCAs, heterocyclics and other antidepressive agents, but considerable structural differences also exist even among the various SSRIs. Therefore, some differential pharmacology between the drugs in the same class may be expected. However, head-to-head comparisons found no evidence to suggest side effect profile differences between fluvoxamine and other SSRIs, except for sweating, which was more common in recipients of paroxetine than fluvoxamine in a single trial.

Overall completeness and applicability of evidence

This review has a number of limitations in overall completeness and applicability.

Participants

Severity of depression was believed to be associated with good treatment response in patients taking TCAs compared with those taking SSRIs (Anderson 1994), However, all of the studies included in this review, other than that of Claghorn 1996, were conducted for moderate/mild depression. For this reason, the findings from this review may not be representative of more severely affected patients.

In addition, the findings from this review may not be representative for the depressive elderly, since only one trial was found in which recruitment was wholly conducted for the depressive elderly (Bocksberger 1993). Elder patients with depression are recognized as being more vulnerable to adverse effects of antidepressants (Schneider 1995). If fluvoxamine is more or less acceptable than other antidepressants, this advantage/disadvantage might be obvious in trials focused on geriatric depression.

Twenty-five out of 54 trials explicitly included or might have included patients with major depressive episodes caused by bipolar disorder. However, we found no heterogeneity in the pooled results.

In many countries, including the UK, many patients with depression are treated solely in primary care (Spijker 2001, Goldberg 1991), and depressive disorder in primary care has a different profile than treated in secondary care (Suh 1997). However, we could find only two articles conducted in primary care settings (Barge-Schaapveld 1995; Moon 1991), and the relevance of the studies included in this review to primary care settings should be limited.

Interventions

Considering the often chronic and recurrence-prone presentation of major depression, long-term or follow-up interventions are often required for optimal treatment of this disorder. However, we could find no studies that examined the long-term efficacy of fluvoxamine for major depression.

Outcomes

Treatments for major depression should be assessed not only by psychiatric symptoms, but also by general functioning and/or QOL. However, no trials included in this review incorporated those outcomes. Considering that major depression is associated with a marked personal, social and economic morbidity, the under-investigation of these outcomes borders on negligence. In addition, no studies included cost to health care services as an outcome. The choice of antidepressants may be influenced by factors such as safety of the medication in overdose and the propensity of the medication to be associated with withdrawal symptoms. However, we found no randomised trials that examined these factors in this review.

Quality of the evidence

Fifty-four trials with 4353 patients were included in this analysis; 2117 were randomised to fluvoxamine and 2236 to comparator drugs. These included trials had a number of methodological shortcomings.

Randomisation

All of the trials except one (Rossini 2005) failed to describe methods of random sequence generation. In addition, no trials reported the method of allocation concealment. Therefore, it is conceivable that selection bias might have occurred in the trials included in this review.

Blinding

Information on blinding was sought for many trials; however, no test of blinding success was conducted in any study. The use of an independent, blind assessor was explicitly described in only three studies (Miller 2001; Otsubo 2005; Rossini 2005). On the whole, little information was presented on the outcome assessment process, and the extent to which detection bias might have occurred was uncertain.

Selective outcome reporting

Forty-four out of the 54 included studies used the HRSD as a primary or secondary outcome measure, while a minority of studies used the MADRS and CGI. It is conceivable that the study authors did not report some outcomes in which results failed to show statistical significance. Therefore, the results of our meta-analyses could overestimate the intervention effect due to outcome reporting bias. In addition, only a few studies reported SDs for a change and endpoint score of any depression scale that we adopted as a secondary outcome, which meant that we had to borrow SDs from other trials that did report SDs for the outcomes.

High discontinuation rate

Twenty-seven of the 54 studies had dropout rates higher than 20%. This high attrition rate could have influenced treatment outcomes; for two studies with high attrition rate (Mullin 1988 and Rahman 1991), sensitivity analysis performing the worst case scenario revealed the statistically significant results in favour of the comparator and the best case scenario in favour of fluvoxamine.

Sample size

If we assumed that response rate for experimental antidepressants was 50% and that of the control was 35%, 183 participants would be needed in each treatment group to detect this type of difference at 80% power and 95% confidence. However, most trials included in this review were small; the majority of the studies (38 RCTs) recruited less than 100 participants in total. This made it difficult to interpret the negative finding. In addition, most trials did not discuss their low power or did not conduct sample size calculation

Potential biases in the review process

Strength

The comprehensiveness of our study search and the strict quality appraisal we required before any study could be included in the final pooling of the results contributed to the strength of our review. As a result, randomised evidence from unpublished (Schoemaker 2002) as well as published studies was included in the review. Furthermore, our strong efforts made to obtain missing data resulted in additional data from investigators from 15 trials.

We also imputed response and remission outcomes by applying a threshold of the standard depression severity scales, such as the HRSD or MADRS, with a validated statistical method if they were not available in original trials (Furukawa 2005: Furukawa 2006). We believe our methodology can be used in future systematic reviews in order to minimise outcome reporting bias and to make the most use of information from the obtained trials.

Limitations

This systematic review is not without methodological problems.

First of all, although neither the funnel plot nor Egger's test detected small-study effects, we still cannot rule out the risk of publication bias. There were only 16 studies included in the funnel plots, and all studies were of similar size; therefore, it was difficult to find a meaningful result from funnel plots and their statistical tests. In addition, we have concerns about those five studies that only reported biochemical or physiological outcomes, instead of reporting clinical efficacy and/or tolerability outcomes. For example, one RCT reported prolactin responses to d-fenfluramine for depressive patients, before and after medication but included no clinical outcome at all (Kavoussi 1999). This trial formed part of an industry drug trial sponsored by Solvay, the primary fluvoxamine marketer. We were unable to locate a trial matching the description in this report, and we strongly suspect that we are missing a large trial sponsored by this company. We excluded a crossover trial that did not report the result of first randomisation period (Emrich 1987) and a trial that only reported the result of a maintenance period (White 1990). We tried to contact the original authors to obtain missing outcomes but our attempts were in vain. We believe that publication bias remained a very real risk in this review.

Second, pharmaceutical companies marketing fluvoxamine sponsored a large majority of the trials comparing fluvoxamine with TCAs; the majority of the authors of these trials set fluvoxamine as an investigational drug. Therefore, comparability between fluvoxamine and TCAs in efficacy and tolerability may not be unconditionally warranted; that is to say, the efficacy and tolerability of fluvoxamine over TCAs might be overestimated (Barbui 2004).

Third, although sensitivity analyses excluding trials for which the response rates had to be calculated based on the imputation method did not find any evidence that it changed the results of the main analyses, this could have happened for many reasons other than validity of the imputation methods. For example, there were few studies included in each analysis and most studies were small-sized trials that reported null effects; therefore, removing some trials is unlikely to have a large effect.

Finally, very few of the trials employed standardized instruments in the reporting of side effects. Many of the side effects experienced by patients taking antidepressants may be confused with symptoms and signs of depression. Many trials reported the number of patients who experienced unwanted symptoms during trials, but some articles defined side effects strictly as experiences that appeared for the first time during the treatment period, or experiences that appeared between screen and baseline, but increased in severity during the treatment period. Some trials did not report a side effect profile at all. Obviously the emphasis on detecting side effects differs between trials, and this may explain some of the observed differences.

Agreements and disagreements with other studies or reviews

Edwards 1999 conducted a well-designed systematic review and meta-analysis of RCTs, involving direct comparisons between five SSRIs in the treatment of major depressive illness. The review reported that significantly more patients on fluvoxamine stopped treatment due to any reason and due to side effects compared with other SSRIs. However, this review included only five trials involving fluvoxamine, and these are outdated. The clinical guidelines released by the same authors (Anderson 2001) suggested that fluvoxamine was not the best SSRI choice in routine practice because of its relatively high discontinuation rate, but the results of our review do not provide any evidence that support or argue against this statement.

We have recently published a multiple-treatment meta-analysis (MTM) in which our data for fluvoxamine were merged with those for 11 other new generation antidepressants and both direct and indirect comparisons among them were statistically pooled (Cipriani 2009). The MTM offers a clinically meaningful synthesis when several competing treatments are available for one disease (Lumley 2002; Lu 2006; Salanti 2008), as is the case with major depression, while examining the overall strength and consistency of this network of evidence. The corresponding ORs and their 95%CI for response and total dropout are tabulated in Table 4.

Table 4. Ratio of ORs from MTM and this review
OutcomesOR (95%CI)Ratio of ORs*
MTM (direct + indirect comparisons) (Cipriani 2009)Direct comparisons (This review)
Response
paroxetine0.96 (0.76 to 1.23)0.83 (0.51 to 1.34)1.16
sertraline0.79 (0.61 to 1.01)1.21 (0.53 to 2.75) 0.65
fluoxetine0.98 (0.77 to 1.23)1.00 (0.62 to 1.61)0.98
citalopram0.88 (0.68 to 1.16)0.90 (0.50 to 1.62)0.98
milnacipran0.97 (0.68 to 1.37)0.57 (0.26 to 1.23) 1.70
venlafaxine0.77 (0.59 to 0.99)0.40 (0.18 to 0.92) 1.93
mirtazapine0.71 (0.55 to 0.92)0.88 (0.59 to 1.31)0.81
Total dropout
paroxetine1.10 (0.84 to 1.47)1.08 (0.63 to 1.85)**1.02
sertraline1.38 (1.03 to 1.89)1.46 (0.19 to 11.16)**0.95
fluoxetine1.22 (0.93 to 1.61)1.17 (0.66 to 2.09)**1.04
citalopram1.37 (1.01 to 1.85)1.42 (0.75 to 2.67)0.96
milnacipran1.18 (0.76 to 1.75)1.22 (0.54 to 2.77)**0.97
venlafaxine1.14 (0.86 to 1.54)2.29 (0.97 to 5.43) 0.50
mirtazapine1.18 (0.87 to 1.61)0.83 (0.52 to 1.33)1.42
MTM, multiple-treatments meta-analysis; OR, odds ratio; CI, confidence interval;
For response, ORs higher than 1 favor fluvoxamine. For total dropout, ORs lower than one favour fluvoxamine;
* ORs of this review as reference;
**Data from three comparisons (Ansseau 1991a, Ansseau 1991b, and Gonul 1999) were omitted because these 4-week trials were not included in MTM (Cipriani 2009).

We note that:
1) all of the confidence intervals overlap widely between direct comparisons and MTM (direct + indirect) comparisons, mainly because the 95%CI for direct comparisons are wide, generally indicating that the network of evidence is consistent.
2) When nominal superiority of fluvoxamine existed in direct comparisons (fluvoxamine over sertraline for response, and fluvoxamine over mirtazapine for dropout), this was lost in the MTM, although again their 95%CI overlapped widely.
3) When the ratio of ORs was greater than 1.5 or smaller than 0.67 (versus sertraline, versus milnacipran and versus venlafaxine for response, and versus venlafaxine for dropout), the MTM results were almost always less extreme (i.e., closer to 1.0) than the direct comparisons, very probably because of cancelling out possible sponsorship, publication and other biases.

The relative merits and demerits of direct versus MTM comparisons are still debatable (Bucher 1997; Song 2003; Ioannidis 2006) and we need to carefully weigh and synthesize the direct with the indirect comparisons.

Authors' conclusions

Implications for practice

The main finding of the present study is that the present review is based on poor quality evidence of primary studies and we found no strong evidence that fluvoxamine was either superior of inferior to any other antidepressants, including TCAs and SSRIs, in terms of efficacy and tolerability during the treatment of depression in the acute phase. On the other hand, evidence is clear for differing side effect profiles, especially in terms of gastrointestinal side effects between fluvoxamine and TCAs. The results of the study led us to conclude that clinicians should focus on practical or clinically relevant considerations, including these differences in side effect profiles.

Implications for research

We could have learnt more about the effects of fluvoxamine if the studies included in the review had clearly described sequence generation, allocation concealment, blinding of outcome assessors, and reasons for dropout. In addition to assessment of depressive symptoms and levels of general functioning, QOL and patient satisfaction would be most informative. If continuous rating scales are to be employed, a concerted effort should be made to come up with an agreement as to which measures are the most useful. For analyses of side effect profiles, more reliable and consistent methods of monitoring and reporting side effects during the trial are needed.

Acknowledgements

This review is one publication of the Meta-Analyses of New Generation Antidepressants (MANGA) project in which a group of researchers within the Cochrane Collaboration Depression, Anxiety and Neurosis Group agreed to conduct a systematic review of all available evidence for 12 new generation antidepressants to inform clinical practice and mental health policies. We would like to thank Julian Higgins and Georgia Salanti for their helpful comments and feedback on this review. We would like to thank the CCDAN Editorial Team for their support, information and advice.

We also wish to thank Dr Tatsuo Akechi for excellent and thorough comments on all aspects of the review and Dr Nicolai Ostrovischi for generously translating a Russian article into English and extracting data for this review.

We sent several letters asking for extra information about their trials to authors. Dr Alan Feiger, Dr Alessandro Serretti, Dr Ari Kiev, Dr Arif Khan, Dr Burl Daviss, Dr Carlos Augusto de Mendonca Lima, Dr Carmen Lefebvre, Dr Charles B. Nemeroff, Dr Christine Ehlerding, Dr Daniela Q.C.M. Barge-Schaapveld, Dr David Hackett , Dr Dorotea Muck-Seler, Dr Doug Keller, Dr Emil Coccaro, Dr Eugene Paykel, Dr Geroge Wan, Dr Jackie Gollan, Dr Jacqueline Strik, Dr James M. Perel, Dr Jean Dalery, Dr Jean-Pierre Gachoud, Dr Joep Schoemaker, Dr Johan Denollet, Dr L. Lee Tynes, Dr Lev Sverdlov, Dr Marc Ansseau, Dr Marco Mula, Dr Mario Amore, Dr Masaki Kato, Dr Matt Lehman, Dr Michael Mueck-Weymann, Dr Michael Poole, Dr Nancy A. Nicolson, Dr Nobuhisa Ueda, Dr R. Michael Poole, Dr Raffaella Zanardi, Dr Ram K. Shrivastava, Dr Reiji Yoshimura, Dr Rob Sival, Dr Robert Ferdinand, Dr Robert N. Golden, Dr Roberto Dominguez, Dr Ronald A. Remick, Dr Sandra Kooij, Dr Siegfried Kasper, Dr Tenpei Otsubo, Dr Thomas Rechlin, Dr Victoria J. Grochocinski, Dr Walter Brown, Dr Walter D. Lawhorn, and Dr Yvon.D. Lapierre, were kind enough to respond, for which we are very grateful.

Data and analyses

Download statistical data

Comparison 1. Fluvoxamine versus TCAs
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Response (acute phase): Primary outcome16935Odds Ratio (M-H, Random, 95% CI)0.97 [0.73, 1.29]
1.1 vs Imipramine6375Odds Ratio (M-H, Random, 95% CI)0.97 [0.59, 1.58]
1.2 vs Clomipramine2129Odds Ratio (M-H, Random, 95% CI)0.84 [0.38, 1.85]
1.3 vs Amitriptyline4185Odds Ratio (M-H, Random, 95% CI)0.79 [0.35, 1.75]
1.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
1.5 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.11 [0.55, 2.24]
1.6 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
2 Response (early phase)252148Odds Ratio (M-H, Random, 95% CI)0.95 [0.78, 1.15]
2.1 vs Imipramine121213Odds Ratio (M-H, Random, 95% CI)1.00 [0.78, 1.29]
2.2 vs Clomipramine5299Odds Ratio (M-H, Random, 95% CI)1.04 [0.63, 1.71]
2.3 vs Amitriptyline4397Odds Ratio (M-H, Random, 95% CI)0.40 [0.13, 1.19]
2.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)1.65 [0.58, 4.73]
2.5 vs Desipramine140Odds Ratio (M-H, Random, 95% CI)0.78 [0.19, 3.13]
2.6 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)0.97 [0.35, 2.68]
3 Remission (early phase)252148Odds Ratio (M-H, Random, 95% CI)0.95 [0.73, 1.22]
3.1 vs Imipramine121213Odds Ratio (M-H, Random, 95% CI)1.00 [0.71, 1.42]
3.2 vs Clomipramine5299Odds Ratio (M-H, Random, 95% CI)0.75 [0.40, 1.41]
3.3 vs Amitriptyline4397Odds Ratio (M-H, Random, 95% CI)0.41 [0.10, 1.76]
3.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)2.33 [0.54, 10.14]
3.5 vs Desipramine140Odds Ratio (M-H, Random, 95% CI)1.0 [0.13, 7.89]
3.6 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)0.97 [0.18, 5.16]
4 Remission (acute phase)16935Odds Ratio (M-H, Random, 95% CI)1.00 [0.69, 1.45]
4.1 vs Imipramine6375Odds Ratio (M-H, Random, 95% CI)1.07 [0.59, 1.94]
4.2 vs Clomipramine2129Odds Ratio (M-H, Random, 95% CI)0.64 [0.28, 1.49]
4.3 vs Amitriptyline4185Odds Ratio (M-H, Random, 95% CI)0.61 [0.28, 1.31]
4.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)1.78 [0.67, 4.77]
4.5 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.06 [0.48, 2.35]
4.6 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.5 [1.31, 15.42]
5 Depression scale - Endpoint score: low=good (early phase)5186Std. Mean Difference (IV, Random, 95% CI)0.28 [-0.01, 0.57]
5.1 vs Imipramine158Std. Mean Difference (IV, Random, 95% CI)0.52 [-0.00, 1.05]
5.2 vs Amitriptyline253Std. Mean Difference (IV, Random, 95% CI)0.47 [-0.08, 1.02]
5.3 vs Desipramine135Std. Mean Difference (IV, Random, 95% CI)0.18 [-0.49, 0.84]
5.4 vs Dothiepin140Std. Mean Difference (IV, Random, 95% CI)-0.21 [-0.83, 0.41]
6 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data  Other dataNo numeric data
6.1 vs Imipramine  Other dataNo numeric data
6.2 vs Clomipramine  Other dataNo numeric data
6.3 vs Amitriptyline  Other dataNo numeric data
6.4 vs Nortriptyline  Other dataNo numeric data
7 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data  Other dataNo numeric data
7.1 vs Imipramine  Other dataNo numeric data
7.2 vs Clomipramine  Other dataNo numeric data
7.3 vs Amitriptyline  Other dataNo numeric data
7.4 vs Nortriptyline  Other dataNo numeric data
7.5 vs Dothiepin  Other dataNo numeric data
7.6 vs Desipramine  Other dataNo numeric data
8 Depression scale - Change score: decrease=good (early phase)2119Std. Mean Difference (IV, Random, 95% CI)0.66 [-0.33, 1.64]
8.1 vs Amitriptyline158Std. Mean Difference (IV, Random, 95% CI)1.17 [0.61, 1.73]
8.2 vs Nortriptyline161Std. Mean Difference (IV, Random, 95% CI)0.16 [-0.34, 0.66]
9 Depression scale - Change score: decrease=good (early phase) - missing SDs  Other dataNo numeric data
9.1 vs Imipramine  Other dataNo numeric data
9.2 vs Clomipramine  Other dataNo numeric data
9.3 vs Amitriptyline  Other dataNo numeric data
9.4 vs Desipramine  Other dataNo numeric data
9.5 vs Dothiepine  Other dataNo numeric data
10 Depression scale - Change score: decrease=good (acute phase)3372Std. Mean Difference (IV, Random, 95% CI)0.28 [-0.22, 0.79]
10.1 vs Imipramine2322Std. Mean Difference (IV, Random, 95% CI)0.17 [-0.46, 0.81]
10.2 vs Amitriptyline150Std. Mean Difference (IV, Random, 95% CI)0.55 [-0.01, 1.12]
11 Depression scale - Change score: decrease=good (acute phase) - missing SDs  Other dataNo numeric data
11.1 vs Imipramine  Other dataNo numeric data
11.2 vs Clomipramine  Other dataNo numeric data
11.3 vs Amitriptyline  Other dataNo numeric data
11.4 vs Desipramine  Other dataNo numeric data
11.5 vs Dothiepine  Other dataNo numeric data
12 Total Dropout282268Odds Ratio (M-H, Random, 95% CI)0.96 [0.79, 1.18]
12.1 vs Imipramine131263Odds Ratio (M-H, Random, 95% CI)1.07 [0.81, 1.41]
12.2 vs Clomipramine5299Odds Ratio (M-H, Random, 95% CI)1.08 [0.58, 2.00]
12.3 vs Amitriptyline5420Odds Ratio (M-H, Random, 95% CI)0.78 [0.51, 1.17]
12.4 vs Nortriptiyline174Odds Ratio (M-H, Random, 95% CI)0.54 [0.21, 1.41]
12.5 vs Desipramine287Odds Ratio (M-H, Random, 95% CI)1.59 [0.24, 10.70]
12.6 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.05 [0.49, 2.25]
13 Dropout due to inefficacy221651Odds Ratio (M-H, Random, 95% CI)1.12 [0.68, 1.83]
13.1 vs Imipramine10841Odds Ratio (M-H, Random, 95% CI)1.48 [0.77, 2.86]
13.2 vs Clomipramine3158Odds Ratio (M-H, Random, 95% CI)0.93 [0.13, 6.69]
13.3 vs Amitriptyline5420Odds Ratio (M-H, Random, 95% CI)0.60 [0.20, 1.74]
13.4 vs Nortriptiyline174Odds Ratio (M-H, Random, 95% CI)2.25 [0.39, 13.12]
13.5 vs Desipramine285Odds Ratio (M-H, Random, 95% CI)0.0 [0.0, 0.0]
13.6 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)0.46 [0.08, 2.67]
14 Dropout due to side effects241772Odds Ratio (M-H, Random, 95% CI)0.79 [0.60, 1.04]
14.1 vs Imipramine11908Odds Ratio (M-H, Random, 95% CI)0.91 [0.63, 1.32]
14.2 vs Clomipramine3158Odds Ratio (M-H, Random, 95% CI)0.70 [0.24, 1.98]
14.3 vs Amitriptyline5420Odds Ratio (M-H, Random, 95% CI)0.59 [0.35, 1.00]
14.4 vs Nortriptiyline174Odds Ratio (M-H, Random, 95% CI)0.40 [0.10, 1.70]
14.5 vs Desipramine287Odds Ratio (M-H, Random, 95% CI)1.0 [0.13, 7.89]
14.6 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.25 [0.47, 3.32]
15 Number of patients experiencing at least one side effect9663Odds Ratio (M-H, Random, 95% CI)0.70 [0.49, 0.98]
15.1 vs Imipramine2136Odds Ratio (M-H, Random, 95% CI)0.55 [0.18, 1.64]
15.2 vs Clomipramine275Odds Ratio (M-H, Random, 95% CI)0.45 [0.14, 1.43]
15.3 vs Amitriptyline3327Odds Ratio (M-H, Random, 95% CI)0.66 [0.42, 1.04]
15.4 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.10 [0.51, 2.37]
16 Subgroup analysis - Response (acute phase) 1. Fluvoxamine dosing - Standard dosage6413Odds Ratio (M-H, Random, 95% CI)0.88 [0.53, 1.46]
16.1 vs Imipramine2200Odds Ratio (M-H, Random, 95% CI)0.92 [0.51, 1.64]
16.2 vs Amitriptyline169Odds Ratio (M-H, Random, 95% CI)0.41 [0.16, 1.09]
16.3 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
16.4 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
16.5 vs Amitriptyline123Odds Ratio (M-H, Random, 95% CI)0.5 [0.09, 2.84]
17 Subgroup analysis - Response (acute phase) 1. Fluvoxamine dosing - High dosage10522Odds Ratio (M-H, Random, 95% CI)1.07 [0.74, 1.55]
17.1 vs Imipramine4175Odds Ratio (M-H, Random, 95% CI)0.95 [0.38, 2.40]
17.2 vs Clomipramine2129Odds Ratio (M-H, Random, 95% CI)0.84 [0.38, 1.85]
17.3 vs Amitriptyline293Odds Ratio (M-H, Random, 95% CI)1.51 [0.57, 4.02]
17.4 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.11 [0.55, 2.24]
18 Subgroup analysis - Response (acute phase) 2. Comparator dosing - Standard dosage4215Odds Ratio (M-H, Random, 95% CI)1.27 [0.64, 2.50]
18.1 vs Clomipramine143Odds Ratio (M-H, Random, 95% CI)0.57 [0.12, 2.74]
18.2 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.11 [0.55, 2.24]
18.3 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
19 Subgroup analysis - Response (acute phase) 2. Comparator dosing - High dosage15892Odds Ratio (M-H, Random, 95% CI)0.99 [0.74, 1.33]
19.1 vs Imipramine6375Odds Ratio (M-H, Random, 95% CI)0.97 [0.59, 1.58]
19.2 vs Clomipramine186Odds Ratio (M-H, Random, 95% CI)0.96 [0.39, 2.39]
19.3 vs Amitriptyline4185Odds Ratio (M-H, Random, 95% CI)0.79 [0.35, 1.75]
19.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
19.5 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.11 [0.55, 2.24]
19.6 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
20 Subgroup analysis - Response (acute phase) 4. Treatment settings - Inpatient5301Odds Ratio (M-H, Random, 95% CI)1.17 [0.71, 1.92]
20.1 vs Imipramine2103Odds Ratio (M-H, Random, 95% CI)1.59 [0.70, 3.61]
20.2 vs Clomipramine186Odds Ratio (M-H, Random, 95% CI)0.96 [0.39, 2.39]
20.3 vs Amitriptyline160Odds Ratio (M-H, Random, 95% CI)1.0 [0.26, 3.89]
20.4 vs Dothiepin152Odds Ratio (M-H, Random, 95% CI)1.0 [0.34, 2.98]
21 Subgroup analysis - Response (acute phase) 4. Treatment settings - Outpatients9564Odds Ratio (M-H, Random, 95% CI)0.84 [0.58, 1.20]
21.1 vs Imipramine4272Odds Ratio (M-H, Random, 95% CI)0.80 [0.46, 1.39]
21.2 vs Clomipramine143Odds Ratio (M-H, Random, 95% CI)0.57 [0.12, 2.74]
21.3 vs Amitriptyline2102Odds Ratio (M-H, Random, 95% CI)0.91 [0.17, 5.00]
21.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
21.5 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)1.19 [0.47, 3.00]
22 Sensitivity analysis - Response (acute phase) 2. Excluding trials dropout rate >20%5185Odds Ratio (M-H, Random, 95% CI)0.78 [0.30, 2.07]
22.1 vs Imipramine272Odds Ratio (M-H, Random, 95% CI)0.48 [0.12, 1.98]
22.2 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
22.3 vs Clomipramine143Odds Ratio (M-H, Random, 95% CI)0.57 [0.12, 2.74]
22.4 vs Amitriptyline123Odds Ratio (M-H, Random, 95% CI)0.5 [0.09, 2.84]
23 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT16935Odds Ratio (M-H, Random, 95% CI)0.57 [0.40, 0.81]
23.1 vs Imipramine6375Odds Ratio (M-H, Random, 95% CI)0.60 [0.39, 0.90]
23.2 vs Clomipramine2129Odds Ratio (M-H, Random, 95% CI)0.84 [0.38, 1.85]
23.3 vs Amitriptyline4185Odds Ratio (M-H, Random, 95% CI)0.34 [0.13, 0.90]
23.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
23.5 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)0.34 [0.16, 0.72]
23.6 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
24 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT16935Odds Ratio (M-H, Random, 95% CI)1.78 [1.18, 2.69]
24.1 vs Imipramine6375Odds Ratio (M-H, Random, 95% CI)1.75 [0.78, 3.92]
24.2 vs Clomipramine2129Odds Ratio (M-H, Random, 95% CI)0.99 [0.44, 2.20]
24.3 vs Amitriptyline4185Odds Ratio (M-H, Random, 95% CI)1.67 [0.82, 3.37]
24.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
24.5 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)4.04 [1.85, 8.81]
24.6 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
25 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response5343Odds Ratio (M-H, Random, 95% CI)0.90 [0.57, 1.41]
25.1 vs Imipramine1100Odds Ratio (M-H, Random, 95% CI)0.92 [0.42, 2.02]
25.2 vs Clomipramine186Odds Ratio (M-H, Random, 95% CI)0.96 [0.39, 2.39]
25.3 vs Amitriptyline283Odds Ratio (M-H, Random, 95% CI)0.77 [0.26, 2.24]
25.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
26 Sensitivity analysis - Mean change from baseline (acute phase) 5. Excluding trials for which the SD had to be borrowed from other trials3148Std. Mean Difference (IV, Random, 95% CI)0.39 [0.07, 0.72]
26.1 vs Imipramine124Std. Mean Difference (IV, Random, 95% CI)0.64 [-0.19, 1.47]
26.2 vs Amitriptyline150Std. Mean Difference (IV, Random, 95% CI)0.55 [-0.01, 1.12]
26.3 vs Nortriptyline174Std. Mean Difference (IV, Random, 95% CI)0.22 [-0.24, 0.67]
27 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as an investigational drug14860Odds Ratio (M-H, Random, 95% CI)0.99 [0.72, 1.36]
27.1 vs Imipramine6375Odds Ratio (M-H, Random, 95% CI)0.97 [0.59, 1.58]
27.2 vs Clomipramine2129Odds Ratio (M-H, Random, 95% CI)0.84 [0.38, 1.85]
27.3 vs Amitriptyline3162Odds Ratio (M-H, Random, 95% CI)0.90 [0.32, 2.50]
27.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
27.5 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)1.19 [0.47, 3.00]
27.6 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
28 Sensitivity analysis - Response (acute phase) 7. Funding - Excluding trials funded by the fluvoxamine marketing company5284Odds Ratio (M-H, Random, 95% CI)1.42 [0.82, 2.43]
28.1 vs Imipramine2103Odds Ratio (M-H, Random, 95% CI)1.59 [0.70, 3.61]
28.2 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
28.3 vs Amitriptyline160Odds Ratio (M-H, Random, 95% CI)1.0 [0.26, 3.89]
28.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
29 Sensitivity analysis - Response (acute phase) 7. Funding - Trials funded by the fluvoxamine marketing company11651Odds Ratio (M-H, Random, 95% CI)0.84 [0.61, 1.16]
29.1 vs Imipramine4272Odds Ratio (M-H, Random, 95% CI)0.80 [0.46, 1.39]
29.2 vs Clomipramine2129Odds Ratio (M-H, Random, 95% CI)0.84 [0.38, 1.85]
29.3 vs Amitriptyline3125Odds Ratio (M-H, Random, 95% CI)0.75 [0.25, 2.29]
29.4 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.11 [0.55, 2.24]
30 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression9540Odds Ratio (M-H, Random, 95% CI)1.09 [0.68, 1.73]
30.1 vs Imipramine4303Odds Ratio (M-H, Random, 95% CI)0.90 [0.44, 1.83]
30.2 vs Amitriptyline3116Odds Ratio (M-H, Random, 95% CI)1.16 [0.50, 2.72]
30.3 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
30.4 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
31 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features16935Odds Ratio (M-H, Random, 95% CI)0.97 [0.73, 1.29]
31.1 vs Imipramine6375Odds Ratio (M-H, Random, 95% CI)0.97 [0.59, 1.58]
31.2 vs Clomipramine2129Odds Ratio (M-H, Random, 95% CI)0.84 [0.38, 1.85]
31.3 vs Amitriptyline4185Odds Ratio (M-H, Random, 95% CI)0.79 [0.35, 1.75]
31.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.91 [0.36, 2.28]
31.5 vs Dothiepin2125Odds Ratio (M-H, Random, 95% CI)1.11 [0.55, 2.24]
31.6 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)4.22 [0.98, 18.13]
Analysis 1.1.

Comparison 1 Fluvoxamine versus TCAs, Outcome 1 Response (acute phase): Primary outcome.

Analysis 1.2.

Comparison 1 Fluvoxamine versus TCAs, Outcome 2 Response (early phase).

Analysis 1.3.

Comparison 1 Fluvoxamine versus TCAs, Outcome 3 Remission (early phase).

Analysis 1.4.

Comparison 1 Fluvoxamine versus TCAs, Outcome 4 Remission (acute phase).

Analysis 1.5.

Comparison 1 Fluvoxamine versus TCAs, Outcome 5 Depression scale - Endpoint score: low=good (early phase).

Analysis 1.6.

Comparison 1 Fluvoxamine versus TCAs, Outcome 6 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Imipramine
Amore 1989HRSD-2122.6missing1525.4missing11 
Asakura 2005HRSD-1710.77.014910.67.1149skewed.
Cassano 1986HRSD-1713.4missing12013.7missing119 
Claghorn 1996HRSD-2121.1missing4421.3missing44 
Fabre 1996HRSD-2119missing4618.3missing48 
Feighner 1989HRSD-1712.5missing2116.6missing27 
Guy 1984HRSD-1714.2missing1713missing16 
Itil 1983HRSD-1612.78.2910.46.814skewed.
Koetsier 2002HRSD-1719.49.72715.58.025skewed.
Lydiard 1989HRSD-1718.6missing1717.2missing18 
March 1990HRSD-1715.9missing1317.1missing15 
vs Clomipramine
Coleman 1982HRSD-178.8missing418.6missing43 
Ottevanger 1995HRSD-1715.18.12013.97.420skewed.
vs Amitriptyline
Harris 1991aHRSD-1717.1missing2910.0missing29 
Murasaki 1998aHRSD-1714.57.88016.17.683skewed.
vs Nortriptyline
Otsubo 2005HRSD-1713.57.73215.76.529skewed.

Analysis 1.7.

Comparison 1 Fluvoxamine versus TCAs, Outcome 7 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Imipramine
Claghorn 1996HRSD-2116missing4415.5missing44 
Fabre 1996HRSD-2114.2missing4614.3missing48 
Feighner 1989HRSD-1710.5missing2116.4missing27 
Guy 1984HRSD-1713.0missing1612.1missing16 
Lydiard 1989HRSD-1712.8missing179.0missing18 
March 1990HRSD-1712.5missing1313.8missing15 
vs Clomipramine
de Wilde 1983HRSD-176.4missing215.1missing21 
Zohar 2003HRSD-1713.4missing4212.3missing42 
vs Amitriptyline
Barge-Schaapveld 1995HRSD-1710.05.2118.68.010skewed.
Harris 1991aHRSD-1710.4missing 6.6missing26 
Kostiukova 2003HRSD-177.4missing306.7missing30 
Remick 1994HRSD-178.77.21310.36.010skewed.
vs Nortriptyline
Otsubo 2005HRSD-1711.98.93612.15.938skewed.
vs Dothiepin
Mullin 1988HRSD-178.3missing268.5missing24 
Rahman 1991MADRS13.5missing1914.4missing21 
vs Desipramine
Tourigny-Rivard 1996HRSD-174.9missing2210.8missing25 
Analysis 1.8.

Comparison 1 Fluvoxamine versus TCAs, Outcome 8 Depression scale - Change score: decrease=good (early phase).

Analysis 1.9.

Comparison 1 Fluvoxamine versus TCAs, Outcome 9 Depression scale - Change score: decrease=good (early phase) - missing SDs.

Depression scale - Change score: decrease=good (early phase) - missing SDs
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Imipramine
Amore 1989HRSD-21-3.1missing15-3.4missing11 
Bramanti 1988HRSD-21-11.1missing28-10.4missing30 
Cassano 1986HRSD-17-12.2missing120-12.2missing119 
Claghorn 1996HRSD-21-5missing44-4.6missing44 
Fabre 1996HRSD-21-8.7missing46-8.2missing48 
Feighner 1989HRSD-17-12.5missing21-10.6missing27 
Guy 1984HRSD-17-10.4missing17-12.1missing16 
Itil 1983HRSD-17-7.6missing9-11.5missing14 
Koetsier 2002HRSD-17-7.6missing27-12.2missing25 
Lydiard 1989HRSD-17-5.9missing17-8.2missing18 
March 1990HRSD-17-9.1missing13-8.4missing15 
vs Clomipramine
Coleman 1982HRSD-17-16.6missing41-16missing43 
de Wilde 1983HRSD-17-6.9missing22-6.5missing21 
Dick 1983HRSD-17-11.7missing15-8.6missing13 
Ottevanger 1995HRSD-17-11.3missing20-11.8missing20 
Zohar 2003HRSD-17-10.8missing42-9.8missing42 
vs Amitriptyline
Barge-Schaapveld 1995HRSD-17-6.6missing12-8.2missing10 
Kostiukova 2003HRSD-17-5.3missing30-8.6missing30 
Murasaki 1998aHRSD-17-8.4missing80-7.5missing83 
Remick 1994HRSD-17-6.6missing13-6.9missing9 
vs Desipramine
Nathan 1990HRSD-17-8.4missing17-9.3missing18 
vs Dothiepine
Mullin 1988HRSD-17-6.5missing26-7.2missing24 
Rahman 1991MADRS-10.2missing19-7.9missing21 
Analysis 1.10.

Comparison 1 Fluvoxamine versus TCAs, Outcome 10 Depression scale - Change score: decrease=good (acute phase).

Analysis 1.11.

Comparison 1 Fluvoxamine versus TCAs, Outcome 11 Depression scale - Change score: decrease=good (acute phase) - missing SDs.

Depression scale - Change score: decrease=good (acute phase) - missing SDs
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Imipramine
Claghorn 1996HRSD-21-10.1missing44-10.4missing44 
Fabre 1996HRSD-21-13.5missing46-12.2missing48 
Feighner 1989HRSD-17-14.5missing21-10.8missing27 
Guy 1984HRSD-17-11.6missing16-13missing16 
Lydiard 1989HRSD-17-11.7missing17-17.4missing18 
March 1990HRSD-17-12.5missing13-11.7missing15 
vs Clomipramine
de Wilde 1983HRSD-17-17missing21-19.1missing21 
Zohar 2003HRSD-17-17.2missing42-18.2missing42 
vs Amitriptyline
Barge-Schaapveld 1995HRSD-17-15.7missing11-15.8missing10 
Kostiukova 2003HRSD-17-18.9missing30-19missing30 
Remick 1994HRSD-17-16.3missing13-14.3missing9 
vs Desipramine
Nathan 1990HRSD-17-19.2missing22-13.5missing25 
vs Dothiepine
Mullin 1988HRSD-17-13missing26-12.7missing24 
Rahman 1991MADRS-22.1missing19-20.5missing21 
Analysis 1.12.

Comparison 1 Fluvoxamine versus TCAs, Outcome 12 Total Dropout.

Analysis 1.13.

Comparison 1 Fluvoxamine versus TCAs, Outcome 13 Dropout due to inefficacy.

Analysis 1.14.

Comparison 1 Fluvoxamine versus TCAs, Outcome 14 Dropout due to side effects.

Analysis 1.15.

Comparison 1 Fluvoxamine versus TCAs, Outcome 15 Number of patients experiencing at least one side effect.

Analysis 1.16.

Comparison 1 Fluvoxamine versus TCAs, Outcome 16 Subgroup analysis - Response (acute phase) 1. Fluvoxamine dosing - Standard dosage.

Analysis 1.17.

Comparison 1 Fluvoxamine versus TCAs, Outcome 17 Subgroup analysis - Response (acute phase) 1. Fluvoxamine dosing - High dosage.

Analysis 1.18.

Comparison 1 Fluvoxamine versus TCAs, Outcome 18 Subgroup analysis - Response (acute phase) 2. Comparator dosing - Standard dosage.

Analysis 1.19.

Comparison 1 Fluvoxamine versus TCAs, Outcome 19 Subgroup analysis - Response (acute phase) 2. Comparator dosing - High dosage.

Analysis 1.20.

Comparison 1 Fluvoxamine versus TCAs, Outcome 20 Subgroup analysis - Response (acute phase) 4. Treatment settings - Inpatient.

Analysis 1.21.

Comparison 1 Fluvoxamine versus TCAs, Outcome 21 Subgroup analysis - Response (acute phase) 4. Treatment settings - Outpatients.

Analysis 1.22.

Comparison 1 Fluvoxamine versus TCAs, Outcome 22 Sensitivity analysis - Response (acute phase) 2. Excluding trials dropout rate >20%.

Analysis 1.23.

Comparison 1 Fluvoxamine versus TCAs, Outcome 23 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT.

Analysis 1.24.

Comparison 1 Fluvoxamine versus TCAs, Outcome 24 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT.

Analysis 1.25.

Comparison 1 Fluvoxamine versus TCAs, Outcome 25 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response.

Analysis 1.26.

Comparison 1 Fluvoxamine versus TCAs, Outcome 26 Sensitivity analysis - Mean change from baseline (acute phase) 5. Excluding trials for which the SD had to be borrowed from other trials.

Analysis 1.27.

Comparison 1 Fluvoxamine versus TCAs, Outcome 27 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as an investigational drug.

Analysis 1.28.

Comparison 1 Fluvoxamine versus TCAs, Outcome 28 Sensitivity analysis - Response (acute phase) 7. Funding - Excluding trials funded by the fluvoxamine marketing company.

Analysis 1.29.

Comparison 1 Fluvoxamine versus TCAs, Outcome 29 Sensitivity analysis - Response (acute phase) 7. Funding - Trials funded by the fluvoxamine marketing company.

Analysis 1.30.

Comparison 1 Fluvoxamine versus TCAs, Outcome 30 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression.

Analysis 1.31.

Comparison 1 Fluvoxamine versus TCAs, Outcome 31 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features.

Comparison 2. Fluvoxamine versus Heterocyclics
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Response (acute phase): Primary outcome2125Odds Ratio (M-H, Random, 95% CI)1.25 [0.55, 2.87]
1.1 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)1.25 [0.55, 2.87]
2 Response (early phase)3122Odds Ratio (M-H, Random, 95% CI)0.93 [0.45, 1.94]
2.1 vs Amineptine140Odds Ratio (M-H, Random, 95% CI)1.0 [0.26, 3.87]
2.2 vs Maprotiline282Odds Ratio (M-H, Random, 95% CI)0.91 [0.38, 2.17]
3 Remission (early phase)3122Odds Ratio (M-H, Random, 95% CI)1.25 [0.50, 3.13]
3.1 vs Amineptine140Odds Ratio (M-H, Random, 95% CI)1.0 [0.13, 7.89]
3.2 vs Maprotiline282Odds Ratio (M-H, Random, 95% CI)1.32 [0.47, 3.69]
4 Remission (acute phase)2125Odds Ratio (M-H, Random, 95% CI)2.02 [0.55, 7.39]
4.1 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)2.02 [0.55, 7.39]
5 Depression scale - Endpoint score: low=good (early phase)280Std. Mean Difference (IV, Random, 95% CI)-0.10 [-0.54, 0.34]
5.1 vs Amineptine139Std. Mean Difference (IV, Random, 95% CI)-0.27 [-0.90, 0.36]
5.2 vs Maprotiline141Std. Mean Difference (IV, Random, 95% CI)0.05 [-0.56, 0.66]
6 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data  Other dataNo numeric data
6.1 vs Maprotiline  Other dataNo numeric data
7 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data  Other dataNo numeric data
7.1 vs Mianserin  Other dataNo numeric data
8 Depression scale - Change score: decrease=good (early phase)140Std. Mean Difference (IV, Random, 95% CI)0.01 [-0.61, 0.63]
8.1 vs Maprotiline140Std. Mean Difference (IV, Random, 95% CI)0.01 [-0.61, 0.63]
9 Depression scale - Change score: decrease=good (early phase) - missing SDs  Other dataNo numeric data
9.1 vs Amineptine  Other dataNo numeric data
9.2 vs Maplotiline  Other dataNo numeric data
9.3 vs Mianserin  Other dataNo numeric data
10 Depression scale - Change score: decrease=good (acute phase) - missing SDs  Other dataNo numeric data
10.3 vs Mianserin  Other dataNo numeric data
11 Total Dropout5247Odds Ratio (M-H, Random, 95% CI)0.58 [0.25, 1.34]
11.1 vs Amineptine140Odds Ratio (M-H, Random, 95% CI)0.63 [0.09, 4.24]
11.2 vs Maprotiline282Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 8.26]
11.3 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)0.54 [0.11, 2.65]
12 Dropout due to inefficacy5247Odds Ratio (M-H, Random, 95% CI)0.44 [0.06, 3.14]
12.1 vs Amineptine140Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 8.26]
12.2 vs Maprotiline282Odds Ratio (M-H, Random, 95% CI)0.0 [0.0, 0.0]
12.3 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)0.53 [0.05, 6.21]
13 Dropout due to side effects5247Odds Ratio (M-H, Random, 95% CI)0.80 [0.33, 1.97]
13.1 vs Amineptine140Odds Ratio (M-H, Random, 95% CI)3.15 [0.12, 82.16]
13.2 vs Maprotiline282Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 8.26]
13.3 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)0.75 [0.20, 2.77]
14 Number of patients experiencing at least one side effect3144Odds Ratio (M-H, Random, 95% CI)1.24 [0.46, 3.31]
14.1 vs Maprotiline282Odds Ratio (M-H, Random, 95% CI)1.01 [0.17, 6.00]
14.2 vs Mianserin162Odds Ratio (M-H, Random, 95% CI)1.77 [0.62, 5.06]
15 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT2125Odds Ratio (M-H, Random, 95% CI)0.76 [0.10, 5.62]
15.1 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)0.76 [0.10, 5.62]
16 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT2125Odds Ratio (M-H, Random, 95% CI)3.04 [1.20, 7.67]
16.1 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)3.04 [1.20, 7.67]
17 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response162Odds Ratio (M-H, Random, 95% CI)2.24 [0.51, 9.91]
17.1 vs Mianserin162Odds Ratio (M-H, Random, 95% CI)2.24 [0.51, 9.91]
18 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as an investigational drug162Odds Ratio (M-H, Random, 95% CI)2.24 [0.51, 9.91]
18.1 vs Mianserin162Odds Ratio (M-H, Random, 95% CI)2.24 [0.51, 9.91]
19 Sensitivity analysis - Response (acute phase) 7. Funding - Trials funded by the fluvoxamine marketing company2125Odds Ratio (M-H, Random, 95% CI)1.25 [0.55, 2.87]
19.1 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)1.25 [0.55, 2.87]
20 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression2125Odds Ratio (M-H, Random, 95% CI)1.25 [0.55, 2.87]
20.1 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)1.25 [0.55, 2.87]
21 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features2125Odds Ratio (M-H, Random, 95% CI)1.25 [0.55, 2.87]
21.1 vs Mianserin2125Odds Ratio (M-H, Random, 95% CI)1.25 [0.55, 2.87]
Analysis 2.1.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 1 Response (acute phase): Primary outcome.

Analysis 2.2.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 2 Response (early phase).

Analysis 2.3.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 3 Remission (early phase).

Analysis 2.4.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 4 Remission (acute phase).

Analysis 2.5.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 5 Depression scale - Endpoint score: low=good (early phase).

Analysis 2.6.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 6 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Maprotiline
Mendonca Lima 1997MADRS13.4missing2014.0missing20 

Analysis 2.7.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 7 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Mianserin
Perez 1990MADRS11.97.52213.67.525skewed.
Analysis 2.8.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 8 Depression scale - Change score: decrease=good (early phase).

Analysis 2.9.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 9 Depression scale - Change score: decrease=good (early phase) - missing SDs.

Depression scale - Change score: decrease=good (early phase) - missing SDs
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Amineptine
Brunner 1994HRSD-17-9.7missing19-8.5missing20 
vs Maplotiline
Kasper 1990HRSD-21-12.5missing21-15.4missing20 
vs Mianserin
Moon 1991MADRS-10missing28-11.5missing21 

Analysis 2.10.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 10 Depression scale - Change score: decrease=good (acute phase) - missing SDs.

Depression scale - Change score: decrease=good (acute phase) - missing SDs
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Mianserin
Moon 1991MADRS-22missing28-24.1missing21 
Perez 1990MADRS-23.9missing22-23.6missing25 
Analysis 2.11.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 11 Total Dropout.

Analysis 2.12.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 12 Dropout due to inefficacy.

Analysis 2.13.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 13 Dropout due to side effects.

Analysis 2.14.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 14 Number of patients experiencing at least one side effect.

Analysis 2.15.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 15 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT.

Analysis 2.16.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 16 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT.

Analysis 2.17.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 17 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response.

Analysis 2.18.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 18 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as an investigational drug.

Analysis 2.19.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 19 Sensitivity analysis - Response (acute phase) 7. Funding - Trials funded by the fluvoxamine marketing company.

Analysis 2.20.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 20 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression.

Analysis 2.21.

Comparison 2 Fluvoxamine versus Heterocyclics, Outcome 21 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features.

Comparison 3. Fluvoxamine versus other SSRIs
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Response (acute phase): Primary outcome8967Odds Ratio (M-H, Random, 95% CI)0.96 [0.74, 1.25]
1.1 vs Paroxetine3281Odds Ratio (M-H, Random, 95% CI)0.83 [0.51, 1.34]
1.2 vs Sertraline2185Odds Ratio (M-H, Random, 95% CI)1.21 [0.53, 2.75]
1.3 vs Fluoxetine2284Odds Ratio (M-H, Random, 95% CI)1.00 [0.62, 1.61]
1.4 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.90 [0.50, 1.62]
2 Response (early phase)8967Odds Ratio (M-H, Random, 95% CI)0.98 [0.68, 1.42]
2.1 vs Paroxetine3281Odds Ratio (M-H, Random, 95% CI)0.71 [0.41, 1.24]
2.2 vs Sertraline2185Odds Ratio (M-H, Random, 95% CI)1.32 [0.58, 2.98]
2.3 vs Fluoxetine2284Odds Ratio (M-H, Random, 95% CI)1.23 [0.45, 3.41]
2.4 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.60 [0.19, 1.90]
3 Remission (early phase)7783Odds Ratio (M-H, Random, 95% CI)0.76 [0.42, 1.38]
3.1 vs Paroxetine3281Odds Ratio (M-H, Random, 95% CI)0.65 [0.25, 1.67]
3.2 vs Sertraline2185Odds Ratio (M-H, Random, 95% CI)0.87 [0.29, 2.55]
3.3 vs Fluoxetine1100Odds Ratio (M-H, Random, 95% CI)0.8 [0.25, 2.57]
3.4 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.99 [0.06, 16.04]
4 Remission (acute phase)8967Odds Ratio (M-H, Random, 95% CI)0.98 [0.71, 1.37]
4.1 vs Paroxetine3281Odds Ratio (M-H, Random, 95% CI)0.77 [0.45, 1.33]
4.2 vs Sertraline2185Odds Ratio (M-H, Random, 95% CI)1.31 [0.48, 3.57]
4.3 vs Fluoxetine2284Odds Ratio (M-H, Random, 95% CI)1.24 [0.74, 2.06]
4.4 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.56 [0.23, 1.34]
5 Depression scale - Endpoint score: low=good (early phase)2177Std. Mean Difference (IV, Random, 95% CI)-0.05 [-0.35, 0.24]
5.1 vs Paroxetine193Std. Mean Difference (IV, Random, 95% CI)-0.07 [-0.47, 0.34]
5.2 vs Sertraline184Std. Mean Difference (IV, Random, 95% CI)-0.04 [-0.46, 0.39]
6 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data  Other dataNo numeric data
6.1 vs Paroxetine  Other dataNo numeric data
6.2 vs Sertraline  Other dataNo numeric data
6.3 vs Citalopram  Other dataNo numeric data
6.4 vs Fluoxetine  Other dataNo numeric data
7 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data  Other dataNo numeric data
7.1 vs Paroxetine  Other dataNo numeric data
7.2 vs Sertraline  Other dataNo numeric data
7.3 vs Fluoxetine  Other dataNo numeric data
7.4 vs Citalopram  Other dataNo numeric data
8 Depression scale - Change score: decrease=good (early phase)180Std. Mean Difference (IV, Random, 95% CI)0.20 [-0.24, 0.64]
8.1 vs Paroxetine180Std. Mean Difference (IV, Random, 95% CI)0.20 [-0.24, 0.64]
9 Depression scale - Change score: decrease=good (early phase) - missing SDs  Other dataNo numeric data
9.1 vs Paroxetine  Other dataNo numeric data
9.2 vs Sertraline  Other dataNo numeric data
9.3 vs Citalopram  Other dataNo numeric data
9.4 vs Fluoxetine  Other dataNo numeric data
10 Depression scale - Change score: decrease=good (acute phase)3230Std. Mean Difference (IV, Random, 95% CI)0.10 [-0.23, 0.42]
10.1 vs Paroxetine2138Std. Mean Difference (IV, Random, 95% CI)0.19 [-0.31, 0.68]
10.2 vs Sertraline192Std. Mean Difference (IV, Random, 95% CI)-0.06 [-0.47, 0.35]
11 Depression scale - Change score: decrease=good (acute phase) - missing SDs  Other dataNo numeric data
11.1 vs Paroxetine  Other dataNo numeric data
11.2 vs Sertraline  Other dataNo numeric data
11.3 vs Citalopram  Other dataNo numeric data
11.4 vs Fluoxetine  Other dataNo numeric data
12 Total Dropout91126Odds Ratio (M-H, Random, 95% CI)1.25 [0.93, 1.69]
12.1 vs Paroxetine4334Odds Ratio (M-H, Random, 95% CI)1.03 [0.62, 1.73]
12.2 vs Sertraline3238Odds Ratio (M-H, Random, 95% CI)1.13 [0.24, 5.37]
12.3 vs Fluoxetine3337Odds Ratio (M-H, Random, 95% CI)1.14 [0.65, 2.00]
12.4 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)1.42 [0.75, 2.67]
13 Dropout due to inefficacy4365Odds Ratio (M-H, Random, 95% CI)1.44 [0.33, 6.33]
13.1 vs Paroxetine2180Odds Ratio (M-H, Random, 95% CI)1.00 [0.14, 6.87]
13.2 vs Sertraline2185Odds Ratio (M-H, Random, 95% CI)5.11 [0.24, 109.17]
14 Dropout due to side effects8942Odds Ratio (M-H, Random, 95% CI)1.19 [0.62, 2.28]
14.1 vs Paroxetine4334Odds Ratio (M-H, Random, 95% CI)0.95 [0.28, 3.26]
14.2 vs Sertraline3238Odds Ratio (M-H, Random, 95% CI)1.29 [0.15, 11.33]
14.3 vs Fluoxetine2153Odds Ratio (M-H, Random, 95% CI)0.86 [0.19, 3.89]
14.4 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)1.80 [0.90, 3.59]
15 Number of patients experiencing at least one side effect5478Odds Ratio (M-H, Random, 95% CI)0.89 [0.53, 1.51]
15.1 vs Paroxetine3281Odds Ratio (M-H, Random, 95% CI)0.95 [0.41, 2.23]
15.2 vs Sertraline197Odds Ratio (M-H, Random, 95% CI)0.70 [0.21, 2.37]
15.3 vs Fluoxetine1100Odds Ratio (M-H, Random, 95% CI)0.5 [0.09, 2.86]
16 Subgroup analysis - Response (acute phase) 1. Fluvoxamine dosing - Standard dosage5542Odds Ratio (M-H, Random, 95% CI)0.92 [0.65, 1.30]
16.1 vs Paroxetine2161Odds Ratio (M-H, Random, 95% CI)0.85 [0.45, 1.62]
16.2 vs Sertraline197Odds Ratio (M-H, Random, 95% CI)0.81 [0.36, 1.83]
16.3 vs Fluoxetine2284Odds Ratio (M-H, Random, 95% CI)1.00 [0.62, 1.61]
17 Subgroup analysis - Response (acute phase) 1. Fluvoxamine dosing - High dosage3425Odds Ratio (M-H, Random, 95% CI)1.02 [0.65, 1.60]
17.1 vs Paroxetine1120Odds Ratio (M-H, Random, 95% CI)0.8 [0.38, 1.66]
17.2 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.90 [0.50, 1.62]
17.3 vs Sertraline188Odds Ratio (M-H, Random, 95% CI)1.88 [0.77, 4.63]
18 Subgroup analysis - Response (acute phase) 2. Comparator dosing - Standard dosage2304Odds Ratio (M-H, Random, 95% CI)0.95 [0.60, 1.50]
18.1 vs Paroxetine1120Odds Ratio (M-H, Random, 95% CI)0.8 [0.38, 1.66]
18.2 vs Fluoxetine1184Odds Ratio (M-H, Random, 95% CI)1.06 [0.59, 1.91]
19 Subgroup analysis - Response (acute phase) 2. Comparator dosing - High dosage6663Odds Ratio (M-H, Random, 95% CI)0.96 [0.70, 1.33]
19.1 vs Paroxetine2161Odds Ratio (M-H, Random, 95% CI)0.85 [0.45, 1.62]
19.2 vs Sertraline2185Odds Ratio (M-H, Random, 95% CI)1.21 [0.53, 2.75]
19.3 vs Fluoxetine1100Odds Ratio (M-H, Random, 95% CI)0.89 [0.40, 2.03]
19.4 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.90 [0.50, 1.62]
20 Sensitivity analysis - Response (acute phase) 2. Excluding trials dropout rate >20%2188Odds Ratio (M-H, Random, 95% CI)1.27 [0.61, 2.62]
20.1 vs Sertraline188Odds Ratio (M-H, Random, 95% CI)1.88 [0.77, 4.63]
20.2 vs Fluoxetine1100Odds Ratio (M-H, Random, 95% CI)0.89 [0.40, 2.03]
21 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT8967Odds Ratio (M-H, Random, 95% CI)0.68 [0.43, 1.05]
21.1 vs Paroxetine3281Odds Ratio (M-H, Random, 95% CI)0.33 [0.09, 1.15]
21.2 vs Sertraline2185Odds Ratio (M-H, Random, 95% CI)0.97 [0.47, 1.99]
21.3 vs Fluoxetine2284Odds Ratio (M-H, Random, 95% CI)0.84 [0.52, 1.35]
21.4 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.90 [0.50, 1.62]
22 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT8967Odds Ratio (M-H, Random, 95% CI)1.33 [1.02, 1.74]
22.1 vs Paroxetine3281Odds Ratio (M-H, Random, 95% CI)1.77 [1.08, 2.92]
22.2 vs Sertraline2185Odds Ratio (M-H, Random, 95% CI)1.44 [0.72, 2.90]
22.3 vs Fluoxetine2284Odds Ratio (M-H, Random, 95% CI)1.28 [0.79, 2.07]
22.4 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.90 [0.50, 1.62]
23 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response4622Odds Ratio (M-H, Random, 95% CI)0.89 [0.64, 1.24]
23.1 vs Paroxetine2221Odds Ratio (M-H, Random, 95% CI)0.75 [0.43, 1.31]
23.2 vs Fluoxetine1184Odds Ratio (M-H, Random, 95% CI)1.06 [0.59, 1.91]
23.3 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.90 [0.50, 1.62]
24 Sensitivity analysis - Mean change from baseline (acute phase) 5. Excluding trials for which the SD had to be borrowed from other trials3230Std. Mean Difference (IV, Random, 95% CI)0.10 [-0.23, 0.42]
24.1 vs Paroxetine2138Std. Mean Difference (IV, Random, 95% CI)0.19 [-0.31, 0.68]
24.2 vs Sertraline192Std. Mean Difference (IV, Random, 95% CI)-0.06 [-0.47, 0.35]
25 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as an investigational drug188Odds Ratio (M-H, Random, 95% CI)1.88 [0.77, 4.63]
25.1 vs Sertraline188Odds Ratio (M-H, Random, 95% CI)1.88 [0.77, 4.63]
26 Sensitivity analysis - Response (acute phase) 7. Funding - Excluding trials funded by the fluvoxamine marketing company3425Odds Ratio (M-H, Random, 95% CI)1.02 [0.65, 1.60]
26.1 vs Paroxetine1120Odds Ratio (M-H, Random, 95% CI)0.8 [0.38, 1.66]
26.2 vs Sertraline188Odds Ratio (M-H, Random, 95% CI)1.88 [0.77, 4.63]
26.3 vs Citalopram1217Odds Ratio (M-H, Random, 95% CI)0.90 [0.50, 1.62]
27 Sensitivity analysis - Response (acute phase) 7. Funding - Trials funded by the fluvoxamine marketing company5542Odds Ratio (M-H, Random, 95% CI)0.92 [0.65, 1.30]
27.1 vs Paroxetine2161Odds Ratio (M-H, Random, 95% CI)0.85 [0.45, 1.62]
27.2 vs Sertraline197Odds Ratio (M-H, Random, 95% CI)0.81 [0.36, 1.83]
27.3 vs Fluoxetine2284Odds Ratio (M-H, Random, 95% CI)1.00 [0.62, 1.61]
28 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression3258Odds Ratio (M-H, Random, 95% CI)0.84 [0.51, 1.39]
28.1 vs Paroxetine2161Odds Ratio (M-H, Random, 95% CI)0.85 [0.45, 1.62]
28.2 vs Sertraline197Odds Ratio (M-H, Random, 95% CI)0.81 [0.36, 1.83]
29 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features6630Odds Ratio (M-H, Random, 95% CI)1.01 [0.73, 1.39]
29.1 vs Paroxetine2161Odds Ratio (M-H, Random, 95% CI)0.85 [0.45, 1.62]
29.2 vs Sertraline2185Odds Ratio (M-H, Random, 95% CI)1.21 [0.53, 2.75]
29.3 vs Fluoxetine2284Odds Ratio (M-H, Random, 95% CI)1.00 [0.62, 1.61]
Analysis 3.1.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 1 Response (acute phase): Primary outcome.

Analysis 3.2.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 2 Response (early phase).

Analysis 3.3.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 3 Remission (early phase).

Analysis 3.4.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 4 Remission (acute phase).

Analysis 3.5.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 5 Depression scale - Endpoint score: low=good (early phase).

Analysis 3.6.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 6 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Paroxetine
Kato 2006HRSD-2117.2missing4113.9missing39 
Kiev 1997HRSD-2117.1missing2917.4missing29 
vs Sertraline
Nemeroff 1995HRSD-2116.8missing4615.8missing46 
vs Citalopram
Haffmans 1996HRSD-1722.3missing10921.2missing108 
vs Fluoxetine
Rapaport 1996HRSD-2115.5missing4714.4missing46 

Analysis 3.7.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 7 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Paroxetine
Ansseau 1994HRSD-2113.27.34313.77.938skewed.
Kato 2006HRSD-219.0missing414.6missing39 
Kiev 1997HRSD-2110.97.32911.57.429skewed.
vs Sertraline
Nemeroff 1995HRSD-2114.07.64612.26.546skewed.
Rossini 2005HRSD-217.612.33911.311.345skewed.
vs Fluoxetine
Dalery 2003HRSD-1710.0missing8611.3missing91 
Rapaport 1996HRSD-219.6missing479.7missing46 
vs Citalopram
Haffmans 1996HRSD-1718.0missing10916.6missing108 
Analysis 3.8.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 8 Depression scale - Change score: decrease=good (early phase).

Analysis 3.9.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 9 Depression scale - Change score: decrease=good (early phase) - missing SDs.

Depression scale - Change score: decrease=good (early phase) - missing SDs
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Paroxetine
Ansseau 1994HRSD-21-7.2missing48-6.2missing45 
Kiev 1997HRSD-21-7.3missing29-7missing29 
vs Sertraline
Nemeroff 1995HRSD-21-7.8missing46-7.4missing46 
Rossini 2005HRSD-21-10.1missing39-7.8missing45 
vs Citalopram
Haffmans 1996HRSD-17-2.2missing109-3.5missing108 
vs Fluoxetine
Rapaport 1996HRSD-21-9.7missing47-11.2missing46 
Analysis 3.10.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 10 Depression scale - Change score: decrease=good (acute phase).

Analysis 3.11.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 11 Depression scale - Change score: decrease=good (acute phase) - missing SDs.

Depression scale - Change score: decrease=good (acute phase) - missing SDs
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Paroxetine
Ansseau 1994HRSD-21-13.3missing43-12.3missing38 
vs Sertraline
Rossini 2005HRSD-21-23.7missing39-18missing45 
vs Citalopram
Haffmans 1996HRSD-17-6.5missing109-8.1missing108 
vs Fluoxetine
Dalery 2003HRSD-17-12.3missing86-10.9missing91 
Rapaport 1996HRSD-21-15.6missing47-15.9missing46 
Analysis 3.12.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 12 Total Dropout.

Analysis 3.13.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 13 Dropout due to inefficacy.

Analysis 3.14.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 14 Dropout due to side effects.

Analysis 3.15.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 15 Number of patients experiencing at least one side effect.

Analysis 3.16.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 16 Subgroup analysis - Response (acute phase) 1. Fluvoxamine dosing - Standard dosage.

Analysis 3.17.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 17 Subgroup analysis - Response (acute phase) 1. Fluvoxamine dosing - High dosage.

Analysis 3.18.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 18 Subgroup analysis - Response (acute phase) 2. Comparator dosing - Standard dosage.

Analysis 3.19.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 19 Subgroup analysis - Response (acute phase) 2. Comparator dosing - High dosage.

Analysis 3.20.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 20 Sensitivity analysis - Response (acute phase) 2. Excluding trials dropout rate >20%.

Analysis 3.21.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 21 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT.

Analysis 3.22.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 22 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT.

Analysis 3.23.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 23 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response.

Analysis 3.24.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 24 Sensitivity analysis - Mean change from baseline (acute phase) 5. Excluding trials for which the SD had to be borrowed from other trials.

Analysis 3.25.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 25 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as an investigational drug.

Analysis 3.26.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 26 Sensitivity analysis - Response (acute phase) 7. Funding - Excluding trials funded by the fluvoxamine marketing company.

Analysis 3.27.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 27 Sensitivity analysis - Response (acute phase) 7. Funding - Trials funded by the fluvoxamine marketing company.

Analysis 3.28.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 28 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression.

Analysis 3.29.

Comparison 3 Fluvoxamine versus other SSRIs, Outcome 29 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features.

Comparison 4. Fluvoxamine versus SNRIs
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Response (acute phase): Primary outcome3224Odds Ratio (M-H, Random, 95% CI)0.48 [0.27, 0.85]
1.1 vs Milnacipran1113Odds Ratio (M-H, Random, 95% CI)0.57 [0.26, 1.23]
1.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)0.40 [0.18, 0.92]
2 Response (early phase)5351Odds Ratio (M-H, Random, 95% CI)0.73 [0.45, 1.19]
2.1 vs Milnacipran3240Odds Ratio (M-H, Random, 95% CI)0.65 [0.37, 1.15]
2.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)1.01 [0.39, 2.63]
3 Remission (early phase)5351Odds Ratio (M-H, Random, 95% CI)0.81 [0.35, 1.86]
3.1 vs Milnacipran3240Odds Ratio (M-H, Random, 95% CI)0.62 [0.22, 1.74]
3.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)1.54 [0.22, 10.74]
4 Remission (acute phase)3224Odds Ratio (M-H, Random, 95% CI)0.61 [0.34, 1.08]
4.1 vs Milnacipran1113Odds Ratio (M-H, Random, 95% CI)0.68 [0.30, 1.51]
4.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)0.54 [0.23, 1.24]
5 Depression scale - Endpoint score: low=good (early phase)4274Std. Mean Difference (IV, Random, 95% CI)0.07 [-0.17, 0.32]
5.1 vs Milnacipran2172Std. Mean Difference (IV, Random, 95% CI)0.00 [-0.30, 0.31]
5.2 vs Venlafaxine2102Std. Mean Difference (IV, Random, 95% CI)0.21 [-0.21, 0.63]
6 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data  Other dataNo numeric data
6.1 vs Milnacipran  Other dataNo numeric data
7 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data  Other dataNo numeric data
7.1 vs Venlafaxine  Other dataNo numeric data
8 Depression scale - Change score: decrease=good (early phase) - missing SDs  Other dataNo numeric data
8.1 vs Milnacipran  Other dataNo numeric data
8.3 vs Venlafaxine  Other dataNo numeric data
9 Depression scale - Change score: decrease=good (acute phase) - missing SDs  Other dataNo numeric data
9.1 vs Milnacipran  Other dataNo numeric data
9.3 vs Venlafaxine  Other dataNo numeric data
10 Total Dropout5351Odds Ratio (M-H, Random, 95% CI)1.57 [0.93, 2.67]
10.1 vs Milnacipran3240Odds Ratio (M-H, Random, 95% CI)1.26 [0.65, 2.45]
10.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)2.29 [0.97, 5.43]
11 Dropout due to inefficacy3240Odds Ratio (M-H, Random, 95% CI)0.86 [0.34, 2.16]
11.1 vs Milnacipran3240Odds Ratio (M-H, Random, 95% CI)0.86 [0.34, 2.16]
12 Dropout due to side effects3240Odds Ratio (M-H, Random, 95% CI)2.38 [0.73, 7.78]
12.1 vs Milnacipran3240Odds Ratio (M-H, Random, 95% CI)2.38 [0.73, 7.78]
13 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT3224Odds Ratio (M-H, Random, 95% CI)0.20 [0.05, 0.76]
13.1 vs Milnacipran1113Odds Ratio (M-H, Random, 95% CI)0.57 [0.26, 1.23]
13.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)0.11 [0.04, 0.29]
14 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT3224Odds Ratio (M-H, Random, 95% CI)1.41 [0.43, 4.66]
14.1 vs Milnacipran1113Odds Ratio (M-H, Random, 95% CI)0.57 [0.26, 1.23]
14.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)2.67 [0.98, 7.26]
15 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response1113Odds Ratio (M-H, Random, 95% CI)0.57 [0.26, 1.23]
15.1 vs Milnacipran1113Odds Ratio (M-H, Random, 95% CI)0.57 [0.26, 1.23]
16 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as a comparator drug3224Odds Ratio (M-H, Random, 95% CI)0.48 [0.27, 0.85]
16.1 vs Milnacipran1113Odds Ratio (M-H, Random, 95% CI)0.57 [0.26, 1.23]
16.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)0.40 [0.18, 0.92]
17 Sensitivity analysis - Response (acute phase) 7. Funding - Excluding trials funded by the fluvoxamine marketing company3224Odds Ratio (M-H, Random, 95% CI)0.48 [0.27, 0.85]
17.1 vs Milnacipran1113Odds Ratio (M-H, Random, 95% CI)0.57 [0.26, 1.23]
17.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)0.40 [0.18, 0.92]
18 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression3224Odds Ratio (M-H, Random, 95% CI)0.48 [0.27, 0.85]
18.1 vs Milnacipran1113Odds Ratio (M-H, Random, 95% CI)0.57 [0.26, 1.23]
18.2 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)0.40 [0.18, 0.92]
19 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features2111Odds Ratio (M-H, Random, 95% CI)0.40 [0.18, 0.92]
19.1 vs Venlafaxine2111Odds Ratio (M-H, Random, 95% CI)0.40 [0.18, 0.92]
Analysis 4.1.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 1 Response (acute phase): Primary outcome.

Analysis 4.2.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 2 Response (early phase).

Analysis 4.3.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 3 Remission (early phase).

Analysis 4.4.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 4 Remission (acute phase).

Analysis 4.5.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 5 Depression scale - Endpoint score: low=good (early phase).

Analysis 4.6.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 6 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Milnacipran
Ansseau 1991bHRSD-2420.88.41920.110.142skewed.

Analysis 4.7.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 7 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Venlafaxine
Hackett 1998aMADRS10.511.1117.66.831skewed.
Hackett 1998bMADRS10.511.11010.59.328skewed.

Analysis 4.8.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 8 Depression scale - Change score: decrease=good (early phase) - missing SDs.

Depression scale - Change score: decrease=good (early phase) - missing SDs
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Milnacipran
Ansseau 1991aHRSD-21-11.7missing19-11.6missing42 
Ansseau 1991bHRSD-21-11.7missing18-12.4missing41 
Clerc 2001HRSD-21-10.9missing56-12.6missing57 
vs Venlafaxine
Hackett 1998aMADRS-8.4missing16-11.7missing36 
Hackett 1998bMADRS-8.4missing15-10missing35 

Analysis 4.9.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 9 Depression scale - Change score: decrease=good (acute phase) - missing SDs.

Depression scale - Change score: decrease=good (acute phase) - missing SDs
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Milnacipran
Clerc 2001HRSD-21-15.9missing56-20.7missing57 
vs Venlafaxine
Hackett 1998aMADRS-21.1missing10-22.2missing28 
Hackett 1998bMADRS-21.1missing11-24.1missing31 
Analysis 4.10.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 10 Total Dropout.

Analysis 4.11.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 11 Dropout due to inefficacy.

Analysis 4.12.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 12 Dropout due to side effects.

Analysis 4.13.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 13 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT.

Analysis 4.14.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 14 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT.

Analysis 4.15.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 15 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response.

Analysis 4.16.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 16 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as a comparator drug.

Analysis 4.17.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 17 Sensitivity analysis - Response (acute phase) 7. Funding - Excluding trials funded by the fluvoxamine marketing company.

Analysis 4.18.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 18 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression.

Analysis 4.19.

Comparison 4 Fluvoxamine versus SNRIs, Outcome 19 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features.

Comparison 5. Fluvoxamine versus newer ADs
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Response (acute phase): Primary outcome1 Odds Ratio (M-H, Random, 95% CI)Subtotals only
1.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.88 [0.59, 1.31]
2 Response (early phase)4 Odds Ratio (M-H, Random, 95% CI)Subtotals only
2.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.72 [0.47, 1.11]
2.2 vs Moclobemide3231Odds Ratio (M-H, Random, 95% CI)0.60 [0.26, 1.36]
3 Remission (early phase)4 Odds Ratio (M-H, Random, 95% CI)Subtotals only
3.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.69 [0.36, 1.30]
3.2 vs Moclobemide3231Odds Ratio (M-H, Random, 95% CI)0.47 [0.16, 1.35]
4 Remission (acute phase)1 Odds Ratio (M-H, Random, 95% CI)Subtotals only
4.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)1.19 [0.81, 1.76]
5 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data  Other dataNo numeric data
5.1 vs Mirtazapine  Other dataNo numeric data
5.2 vs Moclobemide  Other dataNo numeric data
6 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data  Other dataNo numeric data
6.1 vs Mirtazapine  Other dataNo numeric data
6.2 vs Moclobemide  Other dataNo numeric data
7 Depression scale - Change score: decrease=good (early phase)1 Std. Mean Difference (IV, Random, 95% CI)Subtotals only
7.1 vs Mirtazapine1402Std. Mean Difference (IV, Random, 95% CI)0.32 [0.12, 0.51]
8 Depression scale - Change score: decrease=good (early phase) - missing SDs  Other dataNo numeric data
8.2 vs Moclobemide  Other dataNo numeric data
9 Depression scale - Change score: decrease=good (acute phase)1 Std. Mean Difference (IV, Random, 95% CI)Subtotals only
9.1 vs Mirtazapine1402Std. Mean Difference (IV, Random, 95% CI)0.08 [-0.12, 0.28]
10 Total Dropout4 Odds Ratio (M-H, Random, 95% CI)Subtotals only
10.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.83 [0.52, 1.33]
10.2 vs Moclobemide3231Odds Ratio (M-H, Random, 95% CI)1.41 [0.73, 2.71]
11 Dropout due to inefficacy4 Odds Ratio (M-H, Random, 95% CI)Subtotals only
11.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)2.02 [0.60, 6.82]
11.2 vs Moclobemide3231Odds Ratio (M-H, Random, 95% CI)0.62 [0.14, 2.71]
12 Dropout due to side effects4 Odds Ratio (M-H, Random, 95% CI)Subtotals only
12.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.60 [0.31, 1.17]
12.2 vs Moclobemide3231Odds Ratio (M-H, Random, 95% CI)1.51 [0.64, 3.53]
13 Number of patients experiencing at least one side effect3 Odds Ratio (M-H, Random, 95% CI)Subtotals only
13.1 vs Moclobemide3231Odds Ratio (M-H, Random, 95% CI)2.29 [1.35, 3.88]
14 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT1 Odds Ratio (M-H, Random, 95% CI)Subtotals only
14.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.77 [0.51, 1.15]
15 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT1 Odds Ratio (M-H, Random, 95% CI)Subtotals only
15.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.95 [0.64, 1.42]
16 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response1 Odds Ratio (M-H, Random, 95% CI)Subtotals only
16.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.88 [0.59, 1.31]
17 Sensitivity analysis - Mean change from baseline (acute phase) 5. Excluding trials for which the SD had to be borrowed from other trials1 Std. Mean Difference (IV, Random, 95% CI)Subtotals only
17.1 vs Mirtazapine1402Std. Mean Difference (IV, Random, 95% CI)0.08 [-0.12, 0.28]
18 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as a comparator drug1 Odds Ratio (M-H, Random, 95% CI)Subtotals only
18.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.88 [0.59, 1.31]
19 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression1 Odds Ratio (M-H, Random, 95% CI)Subtotals only
19.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.88 [0.59, 1.31]
20 Sensitivity analysis - Response (acute phase) 7. Funding - Excluding trials funded by the fluvoxamine marketing company1 Odds Ratio (M-H, Random, 95% CI)Subtotals only
20.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.88 [0.59, 1.31]
21 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features1 Odds Ratio (M-H, Random, 95% CI)Subtotals only
21.1 vs Mirtazapine1412Odds Ratio (M-H, Random, 95% CI)0.88 [0.59, 1.31]
Analysis 5.1.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 1 Response (acute phase): Primary outcome.

Analysis 5.2.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 2 Response (early phase).

Analysis 5.3.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 3 Remission (early phase).

Analysis 5.4.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 4 Remission (acute phase).

Analysis 5.5.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 5 Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (early phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Mirtazapine
Schoemaker 2002HRSD-1716.5missing20314.8missing199 
vs Moclobemide
Bocksberger 1993MADRS31.59.81920.010.819skewed.

Analysis 5.6.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 6 Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data.

Depression scale - Endpoint score: low=good (acute phase) - missing SDs or skewed data
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Mirtazapine
Schoemaker 2002HRSD-1716.5missing20314.8missing199 
vs Moclobemide
Bocksberger 1993MADRS31.59.81920.010.819skewed.
Analysis 5.7.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 7 Depression scale - Change score: decrease=good (early phase).

Analysis 5.8.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 8 Depression scale - Change score: decrease=good (early phase) - missing SDs.

Depression scale - Change score: decrease=good (early phase) - missing SDs
StudyDepression scaleFluvoxamine; meanSDncomparator; meanSDnNote
vs Moclobemide
Barrelet 1991HRSD-17-10.3missing25-10.1missing26 
Bocksberger 1993NADRS-11.8missing19-21.9missing19 
Bougerol 1992HRSD-17-8.4missing61-9.0missing65 
Analysis 5.9.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 9 Depression scale - Change score: decrease=good (acute phase).

Analysis 5.10.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 10 Total Dropout.

Analysis 5.11.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 11 Dropout due to inefficacy.

Analysis 5.12.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 12 Dropout due to side effects.

Analysis 5.13.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 13 Number of patients experiencing at least one side effect.

Analysis 5.14.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 14 Sensitivity analysis - Response (acute phase) 3. Worst case scenario ITT.

Analysis 5.15.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 15 Sensitivity analysis - Response (acute phase) 4. Best case scenario ITT.

Analysis 5.16.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 16 Sensitivity analysis - Response (acute phase) 5. Excluding trials with imputation methods for calculating response.

Analysis 5.17.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 17 Sensitivity analysis - Mean change from baseline (acute phase) 5. Excluding trials for which the SD had to be borrowed from other trials.

Analysis 5.18.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 18 Sensitivity analysis - Response (acute phase) 6. Wish bias - Fluvoxamine as a comparator drug.

Analysis 5.19.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 19 Sensitivity analysis - Response (acute phase) 8. Excluding trials that might include patients with bipolar depression.

Analysis 5.20.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 20 Sensitivity analysis - Response (acute phase) 7. Funding - Excluding trials funded by the fluvoxamine marketing company.

Analysis 5.21.

Comparison 5 Fluvoxamine versus newer ADs, Outcome 21 Sensitivity analysis - Response (acute phase) 9. Excluding trials that included patients with psychotic features.

Comparison 6. Fluvoxamine versus other conventional psychotropic drugs
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Response (early phase)140Odds Ratio (M-H, Random, 95% CI)0.0 [0.0, 0.0]
1.1 vs Sulpiride140Odds Ratio (M-H, Random, 95% CI)0.0 [0.0, 0.0]
2 Remission (early phase)140Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 8.26]
2.1 vs Sulpiride140Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 8.26]
3 Depression scale - Endpoint score: low=good (early phase)140Std. Mean Difference (IV, Random, 95% CI)0.47 [-0.16, 1.10]
3.1 vs Sulpiride140Std. Mean Difference (IV, Random, 95% CI)0.47 [-0.16, 1.10]
4 Depression scale - Change score: decrease=good (early phase)140Std. Mean Difference (IV, Random, 95% CI)0.52 [-0.11, 1.15]
4.1 vs Sulpiride140Std. Mean Difference (IV, Random, 95% CI)0.52 [-0.11, 1.15]
5 Total Dropout140Odds Ratio (M-H, Random, 95% CI)0.0 [0.0, 0.0]
5.1 vs Sulpiride140Odds Ratio (M-H, Random, 95% CI)0.0 [0.0, 0.0]
Analysis 6.1.

Comparison 6 Fluvoxamine versus other conventional psychotropic drugs, Outcome 1 Response (early phase).

Analysis 6.2.

Comparison 6 Fluvoxamine versus other conventional psychotropic drugs, Outcome 2 Remission (early phase).

Analysis 6.3.

Comparison 6 Fluvoxamine versus other conventional psychotropic drugs, Outcome 3 Depression scale - Endpoint score: low=good (early phase).

Analysis 6.4.

Comparison 6 Fluvoxamine versus other conventional psychotropic drugs, Outcome 4 Depression scale - Change score: decrease=good (early phase).

Analysis 6.5.

Comparison 6 Fluvoxamine versus other conventional psychotropic drugs, Outcome 5 Total Dropout.

Comparison 7. Side effect profile: Fluvoxamine vs TCAs
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Cardiovascular - Hypertension / tachycardia4363Odds Ratio (M-H, Random, 95% CI)1.56 [0.51, 4.78]
1.1 vs Imipramine136Odds Ratio (M-H, Random, 95% CI)1.64 [0.31, 8.68]
1.2 vs Clomipramine132Odds Ratio (M-H, Random, 95% CI)5.0 [0.22, 112.88]
1.3 vs Amitriptyline2295Odds Ratio (M-H, Random, 95% CI)1.04 [0.19, 5.81]
2 Cardiovascular - Hypotension / bradycardia8930Odds Ratio (M-H, Random, 95% CI)0.40 [0.21, 0.79]
2.1 vs Imipramine4560Odds Ratio (M-H, Random, 95% CI)0.24 [0.10, 0.62]
2.2 vs Clomipramine275Odds Ratio (M-H, Random, 95% CI)1.23 [0.38, 4.02]
2.3 vs Amitriptyline2295Odds Ratio (M-H, Random, 95% CI)0.20 [0.03, 1.18]
3 Dermatological - Dermatitis / rash3348Odds Ratio (M-H, Random, 95% CI)0.89 [0.17, 4.61]
3.1 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)0.54 [0.05, 5.99]
3.2 vs Desipramine140Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 8.26]
3.3 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)5.14 [0.24, 110.89]
4 Dermatological - Sweating111248Odds Ratio (M-H, Random, 95% CI)0.49 [0.28, 0.88]
4.1 vs Imipramine7972Odds Ratio (M-H, Random, 95% CI)0.32 [0.16, 0.66]
4.2 vs Clomipramine3216Odds Ratio (M-H, Random, 95% CI)0.88 [0.45, 1.73]
4.3 vs Amitriptyline160Odds Ratio (M-H, Random, 95% CI)2.07 [0.18, 24.15]
5 Gastrointestinal - Increased salivation130Odds Ratio (M-H, Random, 95% CI)9.33 [0.96, 90.94]
5.1 vs Imipramine130Odds Ratio (M-H, Random, 95% CI)9.33 [0.96, 90.94]
6 Gastrointestinal - Dry mouth171736Odds Ratio (M-H, Random, 95% CI)0.27 [0.19, 0.38]
6.1 vs Imipramine91055Odds Ratio (M-H, Random, 95% CI)0.24 [0.16, 0.34]
6.2 vs Clomipramine3216Odds Ratio (M-H, Random, 95% CI)0.43 [0.22, 0.81]
6.3 vs Amitriptyline3318Odds Ratio (M-H, Random, 95% CI)0.11 [0.01, 1.06]
6.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.42 [0.16, 1.06]
6.5 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)0.08 [0.01, 0.70]
7 Gastrointestinal - Oral discomfort / taste disturbance2308Odds Ratio (M-H, Random, 95% CI)0.43 [0.06, 2.98]
7.1 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)0.36 [0.01, 8.85]
7.2 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)0.47 [0.04, 5.45]
8 Gastrointestinal - Vomiting / nausea181805Odds Ratio (M-H, Random, 95% CI)2.35 [1.80, 3.07]
8.1 vs Imipramine91055Odds Ratio (M-H, Random, 95% CI)2.23 [1.59, 3.14]
8.2 vs Clomipramine3216Odds Ratio (M-H, Random, 95% CI)2.13 [1.06, 4.27]
8.3 vs Amitriptyline4387Odds Ratio (M-H, Random, 95% CI)2.86 [1.31, 6.23]
8.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)3.01 [1.00, 9.11]
8.5 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)2.71 [0.90, 8.19]
9 Gastrointestinal - Constipation151666Odds Ratio (M-H, Random, 95% CI)0.53 [0.35, 0.80]
9.1 vs Imipramine81008Odds Ratio (M-H, Random, 95% CI)0.50 [0.27, 0.93]
9.2 vs Clomipramine3216Odds Ratio (M-H, Random, 95% CI)0.55 [0.27, 1.09]
9.3 vs Amitriptyline2295Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 7.25]
9.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.64 [0.25, 1.60]
9.5 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)0.31 [0.03, 3.08]
10 Gastrointestinal - Diarrhoea5518Odds Ratio (M-H, Random, 95% CI)2.98 [1.08, 8.20]
10.1 vs Imipramine2136Odds Ratio (M-H, Random, 95% CI)6.38 [1.27, 32.04]
10.2 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)0.36 [0.01, 8.85]
10.3 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)1.82 [0.66, 5.00]
10.4 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)9.81 [0.51, 189.07]
11 Gastrointestinal - Weight gain4425Odds Ratio (M-H, Random, 95% CI)0.53 [0.25, 1.09]
11.1 vs Imipramine130Odds Ratio (M-H, Random, 95% CI)1.0 [0.23, 4.31]
11.2 vs Clomipramine186Odds Ratio (M-H, Random, 95% CI)0.24 [0.05, 1.22]
11.3 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)3.27 [0.13, 81.01]
11.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.44 [0.16, 1.21]
12 Gastrointestinal - Weight loss4226Odds Ratio (M-H, Random, 95% CI)2.76 [1.20, 6.34]
12.1 vs Imipramine266Odds Ratio (M-H, Random, 95% CI)1.88 [0.53, 6.67]
12.2 vs Clomipramine186Odds Ratio (M-H, Random, 95% CI)7.76 [0.91, 66.05]
12.3 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)2.83 [0.79, 10.21]
13 Gastrointestinal - Increased appetite2335Odds Ratio (M-H, Random, 95% CI)0.77 [0.08, 7.04]
13.1 vs Imipramine1100Odds Ratio (M-H, Random, 95% CI)0.32 [0.03, 3.18]
13.2 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)3.27 [0.13, 81.01]
14 Gastrointestinal - Anorexia61104Odds Ratio (M-H, Random, 95% CI)1.17 [0.54, 2.53]
14.1 vs Imipramine5869Odds Ratio (M-H, Random, 95% CI)1.12 [0.47, 2.68]
14.2 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)2.18 [0.19, 24.38]
15 Neuropsychiatric - Blurred vision5500Odds Ratio (M-H, Random, 95% CI)0.47 [0.23, 0.97]
15.1 vs Clomipramine2118Odds Ratio (M-H, Random, 95% CI)0.38 [0.06, 2.30]
15.2 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)0.26 [0.03, 2.39]
15.3 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.57 [0.21, 1.56]
15.4 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)0.47 [0.04, 5.45]
16 Neuropsychiatric - Dizziness / vertigo / faintness141592Odds Ratio (M-H, Random, 95% CI)0.27 [0.19, 0.38]
16.1 vs Imipramine91055Odds Ratio (M-H, Random, 95% CI)0.24 [0.15, 0.38]
16.2 vs Clomipramine186Odds Ratio (M-H, Random, 95% CI)0.21 [0.05, 0.80]
16.3 vs Amitriptyline2304Odds Ratio (M-H, Random, 95% CI)0.31 [0.11, 0.83]
16.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.22 [0.07, 0.70]
16.5 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)0.75 [0.18, 3.06]
17 Neuropsychiatric - Fatigue / tiredness / asthenia5565Odds Ratio (M-H, Random, 95% CI)0.95 [0.52, 1.73]
17.1 vs Imipramine2200Odds Ratio (M-H, Random, 95% CI)1.06 [0.25, 4.56]
17.2 vs Clomipramine2130Odds Ratio (M-H, Random, 95% CI)0.86 [0.40, 1.85]
17.3 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)1.64 [0.27, 9.98]
18 Neuropsychiatric - Headache161621Odds Ratio (M-H, Random, 95% CI)1.09 [0.73, 1.62]
18.1 vs Imipramine81008Odds Ratio (M-H, Random, 95% CI)1.20 [0.71, 2.01]
18.2 vs Clomipramine132Odds Ratio (M-H, Random, 95% CI)1.5 [0.33, 6.83]
18.3 vs Amitriptyline4387Odds Ratio (M-H, Random, 95% CI)0.27 [0.07, 1.11]
18.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)1.82 [0.66, 5.00]
18.5 vs Desipramine147Odds Ratio (M-H, Random, 95% CI)0.86 [0.26, 2.79]
18.6 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)1.5 [0.24, 9.55]
19 Neuropsychiatric - Tremor121247Odds Ratio (M-H, Random, 95% CI)0.50 [0.33, 0.75]
19.1 vs Imipramine6663Odds Ratio (M-H, Random, 95% CI)0.60 [0.32, 1.13]
19.2 vs Clomipramine3216Odds Ratio (M-H, Random, 95% CI)0.45 [0.15, 1.37]
19.3 vs Amitriptyline2295Odds Ratio (M-H, Random, 95% CI)0.40 [0.08, 1.88]
19.4 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)0.31 [0.03, 3.08]
20 Neuropsychiatric - Involuntary movement other than tremor3499Odds Ratio (M-H, Random, 95% CI)1.06 [0.32, 3.47]
20.1 vs Imipramine2430Odds Ratio (M-H, Random, 95% CI)1.28 [0.36, 4.58]
20.2 vs Amitriptyline169Odds Ratio (M-H, Random, 95% CI)0.31 [0.01, 7.99]
21 Neuropsychiatric - Insomnia91039Odds Ratio (M-H, Random, 95% CI)1.26 [0.82, 1.92]
21.1 vs Imipramine6674Odds Ratio (M-H, Random, 95% CI)1.44 [0.87, 2.39]
21.2 vs Clomipramine2130Odds Ratio (M-H, Random, 95% CI)0.82 [0.37, 1.81]
21.3 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)5.49 [0.26, 115.67]
22 Neuropsychiatric - Sleepiness / drowsiness151585Odds Ratio (M-H, Random, 95% CI)0.94 [0.66, 1.33]
22.1 vs Imipramine81019Odds Ratio (M-H, Random, 95% CI)1.08 [0.80, 1.46]
22.2 vs Clomipramine132Odds Ratio (M-H, Random, 95% CI)1.09 [0.25, 4.71]
22.3 vs Amitriptyline4387Odds Ratio (M-H, Random, 95% CI)0.24 [0.05, 1.21]
22.4 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.56 [0.22, 1.42]
22.5 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)1.55 [0.40, 6.02]
23 Neuropsychiatric - Agitation / anxiety8984Odds Ratio (M-H, Random, 95% CI)1.64 [0.94, 2.84]
23.1 vs Imipramine5644Odds Ratio (M-H, Random, 95% CI)2.24 [1.01, 4.97]
23.2 vs Clomipramine132Odds Ratio (M-H, Random, 95% CI)1.87 [0.15, 22.94]
23.3 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)0.71 [0.19, 2.58]
23.4 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)1.25 [0.31, 5.08]
24 Neuropsychiatric - Manic symptom2565Odds Ratio (M-H, Random, 95% CI)0.86 [0.06, 11.58]
24.1 vs Imipramine1330Odds Ratio (M-H, Random, 95% CI)0.23 [0.03, 2.11]
24.2 vs Amitriptyline1235Odds Ratio (M-H, Random, 95% CI)3.3 [0.34, 32.19]
25 Neuropsychiatric - Completed suicide2118Odds Ratio (M-H, Random, 95% CI)2.87 [0.29, 28.65]
25.1 vs Clomipramine2118Odds Ratio (M-H, Random, 95% CI)2.87 [0.29, 28.65]
26 Neuropsychiatric - Suicide wishes / gestures / attempts3489Odds Ratio (M-H, Random, 95% CI)0.65 [0.17, 2.48]
26.1 vs Imipramine1330Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 7.80]
26.2 vs Clomipramine186Odds Ratio (M-H, Random, 95% CI)0.95 [0.18, 5.00]
26.3 vs Dothiepin173Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 8.00]
27 Genitourinary - Problems urinating6818Odds Ratio (M-H, Random, 95% CI)0.44 [0.23, 0.83]
27.1 vs Imipramine2409Odds Ratio (M-H, Random, 95% CI)0.18 [0.04, 0.71]
27.2 vs Amitriptyline2295Odds Ratio (M-H, Random, 95% CI)0.55 [0.10, 2.94]
27.3 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.51 [0.19, 1.38]
27.4 vs Desipramine140Odds Ratio (M-H, Random, 95% CI)0.32 [0.01, 8.26]
28 Genitourinary - Sexual dysfunction174Odds Ratio (M-H, Random, 95% CI)0.75 [0.29, 1.99]
28.1 vs Nortriptyline174Odds Ratio (M-H, Random, 95% CI)0.75 [0.29, 1.99]
Analysis 7.1.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 1 Cardiovascular - Hypertension / tachycardia.

Analysis 7.2.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 2 Cardiovascular - Hypotension / bradycardia.

Analysis 7.3.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 3 Dermatological - Dermatitis / rash.

Analysis 7.4.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 4 Dermatological - Sweating.

Analysis 7.5.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 5 Gastrointestinal - Increased salivation.

Analysis 7.6.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 6 Gastrointestinal - Dry mouth.

Analysis 7.7.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 7 Gastrointestinal - Oral discomfort / taste disturbance.

Analysis 7.8.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 8 Gastrointestinal - Vomiting / nausea.

Analysis 7.9.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 9 Gastrointestinal - Constipation.

Analysis 7.10.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 10 Gastrointestinal - Diarrhoea.

Analysis 7.11.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 11 Gastrointestinal - Weight gain.

Analysis 7.12.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 12 Gastrointestinal - Weight loss.

Analysis 7.13.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 13 Gastrointestinal - Increased appetite.

Analysis 7.14.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 14 Gastrointestinal - Anorexia.

Analysis 7.15.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 15 Neuropsychiatric - Blurred vision.

Analysis 7.16.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 16 Neuropsychiatric - Dizziness / vertigo / faintness.

Analysis 7.17.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 17 Neuropsychiatric - Fatigue / tiredness / asthenia.

Analysis 7.18.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 18 Neuropsychiatric - Headache.

Analysis 7.19.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 19 Neuropsychiatric - Tremor.

Analysis 7.20.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 20 Neuropsychiatric - Involuntary movement other than tremor.

Analysis 7.21.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 21 Neuropsychiatric - Insomnia.

Analysis 7.22.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 22 Neuropsychiatric - Sleepiness / drowsiness.

Analysis 7.23.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 23 Neuropsychiatric - Agitation / anxiety.

Analysis 7.24.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 24 Neuropsychiatric - Manic symptom.

Analysis 7.25.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 25 Neuropsychiatric - Completed suicide.

Analysis 7.26.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 26 Neuropsychiatric - Suicide wishes / gestures / attempts.

Analysis 7.27.

Comparison 7 Side effect profile: Fluvoxamine vs TCAs, Outcome 27 Genitourinary - Problems urinating.