Evidence Table 12Adverse Events

STUDY:Authors: Acharya N et al.179
Year: 2006
Country:
FUNDING:Eli Lilly&Company (A.R., D.N.D., D.G.P., J.P., N.A., and P.C.) and by the Bruce J. Anderson Foundation and the McLean Private Donors Psychopharmacology Research Fund (R.J.B.)
DESIGN:Study design: Pooled data analysis
Number of patients: 2,996
AIMS OF REVIEW:To compare the incidence of suicide-related events with duloxetine versus placebo in controlled trials.
STUDIES INCLUDED IN REVIEW12 placebo-controlled duloxetine trials
TIME PERIOD COVERED:Through February 2, 2004
CHARACTERISTICS OF INCLUDED STUDIES:Double-blind RCTs comparing duloxetine and placebo
CHARACTERISTICS OF INCLUDED POPULATIONS:Adults with MDD
CHARACTERISTICS OF INTERVENTIONS:Duloxetine vs. placebo
MAIN RESULTS:
  • No significant differences in incidence of suicide-related events
  • MHID for suicide-related behaviors was −0.03% (95% CI: −0.48, 0.42) and MHRD −0.002 (95% CI: −0.02, 0.02)
  • Changes in HAM-D Item-3 suicidality scores showed more improvement with duloxetine (MHID, 9.56%; 95% CI: 4.50, 14.6; p < 0.001) and less worsening of suicidal ideation with duloxetine (MHID, −4.25%; 95% CI: −6.55, −1.95; p < 0.001)
  • Other Item-3 findings showed no consistent pattern
  • Analysis found no evidence of increased risk of suicidal behaviors or ideation during treatment with duloxetine vs. placebo in MDD patients
ADVERSE EVENTS:See Main Results
COMPREHENSIVE LITERATURE SEARCH STRATEGY:All completed duloxetine trials in MDD with data lock by February 2, 2004 that were sponsored by the manufacturer, Eli Lilly and Company (16 trials) and by Shionogi Company, Ltd, (11 trials) who hold the license for the development of duloxetine in Japan.
STANDARD METHOD OF APPRAISAL OF STUDIES:NR
QUALITY RATING:Fair
STUDY:Authors: Alper K et al.180
Year: 2007
Country: USA
FUNDING:None
DESIGN:Study design: Retrospective analysis
Setting: FDA reports
Sample size: 38,684 on second-generation antidepressants
INTERVENTION:Citalopram Fluoxetine Venlafaxine Bupropion Paroxetine Nefazodone Mirtazapine Escitalopram
Drug:Duloxetine Sertraline Fluvoxamine
Dose:Various
Duration:1985–2004
Sample size:38,684
INCLUSION:All available public domain data in the form of SBA reports which provided information regarding seizure incidence in phase II and phase III clinical trials. The data set included all of the second-generation antidepressants and atypical antipsychotics
EXCLUSION:Any first generation antipsychotics, or first generation antidepressants except for clomipramine, due to the absence of systematic reporting on seizure incidence in clinical trials for psychotropic drugs approved prior to 1985.
OTHER MEDICATIONS/ INTERVENTIONS:NA
POPULATION CHARACTERISTICS:Groups similar at baseline: NR
Mean age: NR
Gender (female %): NR
Ethnicity: NR
Other population characteristics: NR
OUTCOME ASSESSMENT:Primary Outcome Measures: seizures
Timing of assessments: during RCTs
RESULTS:Incidence of seizure
Bupropion IR0.6%
Citalopram0.3%
Fluoxetine0.2%
Venlafaxine0.1%
Bupropion0.1%
Paroxetine0.07%
Nefazodone0.04%
Mirtazapine0.04%
Escitalopram0%
Duloxetine0%
Sertraline0%
Fluoxetine0.1%
Sertraline0.3%
Fluvoxamine0.2%
ANALYSIS:ITT: NA
Post randomization exclusions: NA
Loss to follow-up: NA
ATTRITION:NA
Withdrawals due to adverse events:
Withdrawals due to lack of efficacy:
Loss to follow-up differential high:
ADVERSE EVENTS:
  • See results
QUALITY RATING:Good
STUDY:Authors: Andersohn et al.181
Year: 2009
Country: United Kingdom
FUNDING:Bayer Schering Pharma AG
DESIGN:Study design: Case control study
Setting: Multi sites – General Practices
Sample size: 165,958
INTERVENTION:Case: Diabetes mellitusControl
Drug:VariousVarious
Dose:VariousVarious
Duration:2 years2 years
Sample size:22438963
INCLUSION:
  • 30 years of age (more likely type 2 diabetes) at the time of cohort entry
  • To be included as a case subject (potential cases of diabetes), a patient had to have at least one prescription of an antidiabetic drug, or two diagnoses of diabetes on different calendar days, or a diagnosis of diabetes and a diabetes-specific test (i.e., glycosylated hemoglobin) on different calendar days. Cohort entry was defined as the date of the first description of an antidepressant
EXCLUSION:
  • The case group: patients who had a suspected diagnosis of diabetes that was not confirmed later on (internal validation)
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 56
Gender (female %): 60.1
Ethnicity (Caucasian %): NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Diabetes
Secondary Outcome Measures: NA
Timing of assessments: NA
RESULTS:Recent long-term use of antidepressants in moderate or high daily doses was associated with an increased risk of diabetes (incidence rate ratio: 1.84; 95% CI, 1.35–2.52).
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS:NA
QUALITY RATING:Fair
STUDY:Authors: Aursnes I, et al.182
Year: 2005
Country: Multinational
FUNDING:NR
DESIGN:Study design: Pooled data analysis
Number of patients: 1,466
AIMS OF REVIEW:To include unpublished data from paroxetine trials for analysis of suicide attempts
STUDIES INCLUDED IN REVIEW16 studies with unpublished data
TIME PERIOD COVERED:NR
CHARACTERISTICS OF INCLUDED STUDIES:Clinical data on paroxetine as presented to world’s drug regulatory agencies in 1989; all double blind, parallel design studies with adult patients randomized to either paroxetine or placebo
CHARACTERISTICS OF INCLUDED POPULATIONS:Adults; patients were excluded from the studies after a suicide-related event
CHARACTERISTICS OF INTERVENTIONS:Paroxetine (no dosage given) vs. placebo
MAIN RESULTS:
  • No suicides in paroxetine or placebo patients
  • 7 suicide attempts in patients on paroxetine and 1 in patients on placebo
  • Probability of increased intensity of suicide attempts per year in adults taking paroxetine was 0.90 with a “pessimistic” prior; probability was somewhat less with 2 more neutral priors
ADVERSE EVENTS:NR
COMPREHENSIVE LITERATURE SEARCH STRATEGY:Yes
STANDARD METHOD OF APPRAISAL OF STUDIES:NR
QUALITY RATING:Fair
STUDY:Authors: Barbui et al.183
Year: 2009
Country: Italy
FUNDING:Fondazione Cariverona, which provided a 3-year grant to the WHO Collaborating Centre for Research and Training in Mental Health and Service Organization at the University of Verona
DESIGN:Study design: Case-control study
Setting: ARNO database, a population-oriented database for drug use in Italy
Sample size: 35,869
INTERVENTION:CasesControlsCasesControls
Drug:Any bleeding disorderAny bleeding disorderGI bleeding disorderGI bleeding disorder
Dose:NANANANA
Duration:NANANANA
Sample size:11,02521,8461,0081,990
INCLUSION:patients admitted between January 1, 2003 and December 31, 2005 whose conditions were diagnosed with abnormal bleeding
EXCLUSION:Prescribed NSAIDs, corticosteroids, antihemorrhagics, and antithrombotic agents
OTHER MEDICATIONS/ INTERVENTIONS:NR except for exclusion
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: NR
Gender (female %): Any 77% and GI 54%
Ethnicity (Caucasian %): NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Any bleeding disorder
Secondary Outcome Measures: GI bleeding
Timing of assessments: NA
RESULTS:
Any bleeding disorderGI bleeding
7% of the cases were exposed to anti-depressants8,6% of the cases were exposed to anti-depressants
6,9% of the controls were exposed to anti-depressants6,3% of the controls were exposed to anti-depressants
Adjusted ORAdjusted OR
No use 1 (reference)1 (reference)
SSRIs 0.99 (95% CI 0.89 to 1.10)1.31 (95% CI 0.91 to 1.88)
Citalopram 0.98 (95% CI 0.79 to 1.20)1.48 (95% CI 0.77 to 2.82)
Escitalopram 1.23 (95% CI 0.84 to 1.81)1.36 (95% CI 0.40 to 4.58)
Fluoxetine 1.01 (95% CI 0.77 to 1.34)0.73 (95% CI 0.19 to 2.79)
Fluvoxamine 0.57 (95% CI 0.29 to 1.13)NA
Paroxetine 0.92 (95% CI 0.78 to 1.10)1.35 (95% CI 0.77 to 2.37)
Sertraline 1.09 (95% CI 0.87 to 1.37)1.77 (95% CI 0.81 to 3.91)
Mirtazapine 0.91 (95% CI 0.63 to 1.32);2.66 (95% CI 0.80 to 8.88);
Venlafaxine 1.07 (95% CI 0.83 to 1.39);1.53 (95% CI 0.60 to 3.91);
Any AD 0.99 (95% CI 0.90 to 1.08) 1.34 (95% CI 1.01 to 1.80)
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS: see results
QUALITY RATING:Good
STUDY:Authors: Barbui et al.184
Year: 2009
Country: NA
FUNDING:grant by Fondazione Cariverona (foundation)
DESIGN:Study design: SR & Meta-Analysis of observational studies
Number of patients: >200,000
AIMS OF REVIEW:To review systematically the risk of attempted and completed suicide after exposure to SSRIs compared to those not exposed to SSRIs in patients with moderate to severe MDD
STUDIES INCLUDED IN REVIEW8 observational studies
TIME PERIOD COVERED:January 1990 to June 2008
CHARACTERISTICS OF INCLUDED STUDIES:6 cohort studies, 2 case-control studies; only studies reporting data on completed or attempted suicide (using ICD-9 or ICD-10 for outcome definition) and compared SSRI use with no use of antidepressants, and where a formal diagnosis or proxy measure of MDD was used, and data in relative risk estimates for re-analysis was reported;
CHARACTERISTICS OF INCLUDED POPULATIONS:Any age, both sexes; diagnosis of MDD (formal diagnosis or proxy measure of MDD)
CHARACTERISTICS OF INTERVENTIONS:SSRI use (citalopram, fluoxetine, fluvoxamine, paroxetine, sertraline, venlafaxine individually and as a class) vs. no-use
MAIN RESULTS:SSRIs as a class vs. no use (pooled, overall risk of completed or attempted suicide, all random effect OR):
  • adolescents: (OR) 1.92, 95%CI (1.51–2.44)
  • adults (OR) 0.57, 95% CI (0.47–0.70)
  • elderly people (>= 65 years) (OR) 0.46, 95% CI (0.27–0.79
Individual SSRIs (data of 2 studies available for each age group):
ADVERSE EVENTS:Completed or attempted suicide (using ICD-9 or ICD-10 for outcome definition (including self-inflicted injury from poisoning, hanging, submersion, firearms, cutting or piercing, jumping from high places, or other means) -> see results for detail
COMPREHENSIVE LITERATURE SEARCH STRATEGY:Yes (see comments)
STANDARD METHOD OF APPRAISAL OF STUDIES:Yes (criteria not reported in article but information given in appendix)
QUALITY RATING:Good
STUDY:Authors: Benkert O, et al.9
Year: 2000
Country: Germany
FUNDING:Organon, GmBH, Munich, Germany
DESIGN:Study design: RCT
Setting: Multi-center (50 centers)
Sample size: 275
INTERVENTION:
Drug:MirtazapineParoxetine
Dose:15–45 mg/d20–40 mg/d
Duration:6 weeks6 weeks
INCLUSION:18–70 years of age; DSM-IV criteria for major depression; ≥ 18 on HAM-D-17
EXCLUSION:Depressive episode longer than 12 months; other psychiatric or psychotic disorder; alcohol or substance abuse; suicidal risk; significant physical illness; non-responders to antidepressants; recent medication with similar drugs; pregnancy
OTHER MEDICATIONS/ INTERVENTIONS:Chloral hydrate for sleep
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: Mirtazapine: 47.2, paroxetine: 47.3
Gender (% female): Mirtazapine: 63%, paroxetine: 65%
Ethnicity: Not reported
Other population characteristics: Not reported
OUTCOME ASSESSMENT:Measures: HAM-D-17, HAM-A, CGI-S, CGI-I, BDI-II, Welzel-Kohnen Colored Scales, Short Form 36
Timing of assessments: Screening, baseline, weeks 1, 2, 3, 4, 6
RESULTS:
  • Mirtazapine and paroxetine were equally effective in reducing mean HAM-D-17 score (58.3% vs. 53.7%)
  • Significantly more mirtazapine patients responded at weeks 1 & 4 on the HAM-D-17 than paroxetine patients; week 1 response: mirtazapine: 23.2%, paroxetine: 8.9% (p < 0.002).
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes
ATTRITION:Loss to follow-up: 23%; mirtazapine: 21.6%, paroxetine: 24.2%
Withdrawals due to adverse events: 8%; mirtazapine: 8.6%, paroxetine: 7.4%
Loss to follow-up differential high: No
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Brambilla P, et al.185
Year: 2005
Country: Multinational
FUNDING:NR
DESIGN:Study design: Meta-analysis
Number of patients: 15,920
AIMS OF REVIEW:To assess the frequency of side-effects in fluoxetine compared to other SSRIs, TCAs and other anti-depressants
STUDIES INCLUDED IN META- ANALYSIS131 studies
TIME PERIOD COVERED:Not reported
CHARACTERISTICS OF INCLUDED STUDIES:All studies with random assigned patients that received fluoxetine or any other anti-depressant. Cross-over studies and those with patients with concomitant medical illness were excluded.
CHARACTERISTICS OF INCLUDED POPULATIONS:Patients with MDD
CHARACTERISTICS OF INTERVENTIONS:Fluoxetine vs. TCA (65 studies); fluoxetine vs. SSRI (22 studies); fluoxetine vs. another AD (44 studies)
MAIN RESULTS:
  • Fluoxetine less withdrawals due to side effects than TCAs and other related Ads RR 0.61 95%CI 0.52, 0.71 but not in comparison to other SSRIs RR 1.04 95% CI 0.84, 1.29
  • Fluoxetine less side effects (50.9%) than TCAs (60.3%) RR= 0.84 95% CI 0.76 to 0.94(p = 0.03) but not in comparison to other SSRIs RR 1.00 95% CI 0.95, 1.04
  • Fluoxetine patients had more activating and GI adverse effects and less cholinergic side effects than other ADs
ADVERSE EVENTS:N/A
COMPREHENSIVE LITERATURE SEARCH STRATEGY:Yes
STANDARD METHOD OF APPRAISAL OF STUDIES:Yes
QUALITY RATING:Good
STUDY:Authors: Bridge JA et al.186
Year: 2007
Country: Multinational
FUNDING:NIMH
DESIGN:Study design: Systematic review and meta-analysis
Number of patients: 5310
AIMS OF REVIEW:To assess the efficacy and risk of reported suicidal ideation/suicide attempt of antidepressants for treatment of pediatric major depressive disorder (MDD), obsessive-compulsive disorder (OCD), and non-OCD anxiety disorders
STUDIES INCLUDED IN REVIEWTwenty-seven trials of pediatric MDD (n = 15), OCD (n = 6), and non-OCD anxiety disorders (n = 6)
TIME PERIOD COVERED:1988 to July 2006
CHARACTERISTICS OF INCLUDED STUDIES:Published and unpublished randomized, placebo-controlled, parallel-group trials of second-generation antidepressants
CHARACTERISTICS OF INCLUDED POPULATIONS:Participants younger than 19 years with MDD, OCD, or non-OCD anxiety disorders
CHARACTERISTICS OF INTERVENTIONS:Second-generation antidepressants (selective serotonin reuptake inhibitors, nefazodone, venlafaxine, and mirtazapine)
MAIN RESULTS:Responder MDD(11.0%; [95% CI, 7.1% to 14.9%]), NNT = 10 (7 to 15)
OCD(19.8% [95% CI, 13.0% to 26.6%), NNT 6 (4 to 8)
Non-OCD anxiety disorders (37.1% [22.5% to 51.7%]), NNT = 3 (2 to 5),
ADVERSE EVENTS:Risk difference of suicidal ideation/suicide attempt across all trials and indications for drug vs placebo (0.7%; 95%CI, 0.1% to 1.3%)
(number needed to harm, 143 [95% CI, 77 to 1000]),
MDD 0.9% (95% CI, −0.1% to 1.9%)
OCD 0.5% (−1.2% to 2.2%)
Non-OCD 0.7% (−0.4% to 1.8%).
Risk difference (95% CI) of Rate of Suicidal Ideation or Suicide Attempt/Preparatory Actions from placebo
MDD
Fluoxetine 2 (−3 to 6)
Paroxetine 2 (−1 to 4)
Escitalopram/citalopram −0 (−3 to 2)
Venlafaxine 4 (1 to 8)
Nefazadone 0 (−1 to 1)
Mirtazapine 1 (−2 to 3)
OCDNon-OCD
Fluoxetine 1 (−4 to 6)Fluoxetine 0 (−5 to 5)
Fluvoxamine 4 (−2 to 9)Fluvoxamine 0(−3 to 3)
Paroxetine 1 (−2 to 4)Paroxetine 2 (−1 to 4)
Sertraline −1 (−4 to 2)Sertraline 0 (−16 to 16)
Venlafaxine 1 (−1 to 2)
COMPREHENSIVE LITERATURE SEARCH STRATEGY:Yes-PubMed (1988 to July 2006), relevant US and British regulatory agency reports, published abstracts of important scientific meetings (1998–2006), clinical trial registries, and information from authors.
STANDARD METHOD OF APPRAISAL OF STUDIES:Yes-according to the criteria of Detsky et al, with final quality ratings based on consensus (intraclass correlation coefficient between raters, 0.94; 95% confidence interval [CI], 0.92 to 0.95)
QUALITY RATING:Good
STUDY:Authors: Buckley NA, et al. 187
Year: 2002
Country: UK
FUNDING:None
DESIGN:Study design: Retrospective database analysis
Setting: General practice
Sample size: 121,927
INTERVENTION:
Drug:TCAs and related drugsSerotoninergic drugs
Dose:VariedVaried
Duration:N/AN/A
Sample size:74,59847,329
INCLUSION:Used TCAs or SSRIs
EXCLUSION:N/A
OTHER MEDICATIONS/INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: N/A
Mean age: NR
Gender (% female): NR
Ethnicity: NR
Other population characteristics: NR
OUTCOME ASSESSMENT:Primary Outcome Measures: Death due to acute poisoning by a single drug w/or w/o co-ingestion of alcohol
Timing of assessments:
RESULTS:

Among second generation antidepressants, venlafaxine had the highest fatal toxicity index (deaths/million prescriptions):

ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Withdrawals due to lack of efficacy: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:
  • See above
QUALITY RATING:N/A
STUDY:Authors: Chen et al. 188
Year: 2008
Country: United States
FUNDING:NR
DESIGN:Study design: Nested case control study
Setting: multi-state managed care organization medical claims data (PHARMetrics).
Sample size: Cohort 587 460 subjects - 1086 cases, 6515 controls (matched by age, sex, and the year of index date of depression)
INTERVENTION:CasesControls
Drug:SSRIs, TCAs and other antidepressants (the other antidepressants were mainly bupropion, phenelzine, tranylcypromine, trazodone nefazodone, venlafaxine)SSRIs, TCAs and other antidepressants (the other antidepressants were mainly bupropion, phenelzine, tranylcypromine, trazodone, nefazodone, venlafaxine)
Dose:NANA
Duration:The window for medication exposure was defined as the time period from the start of the study period to either the index date of a cerebrovascular event, the end of the study period or the end of enrollment, whichever came firstThe window for medication exposure was defined as the time period from the start of the study period to either the index date of a cerebrovascular event, the end of the study period or the end of enrollment, whichever came first
Sample size:10866515
INCLUSION:Non-medicaid patients with depression and at least 6 months continuous enrollment
EXCLUSION:Medicaid recipient
OTHER MEDICATIONS/INTERVENTIONS:Yes
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 34 years or less than 34 years: 5.1%, 35–49 years: 23.7%, 50–64 years: 42.7%, 65–79 years: 17.5% and 80 years or more: 11.1%
Gender (female %): 63.3%
Ethnicity (Caucasian %): NR
Other population characteristics: NR
OUTCOME ASSESSMENT:Primary Outcome Measures: Cerebrovascular Events
Secondary Outcome Measures: NR
Timing of assessments: NA
RESULTS:Risk of a cerebrovascular event compared with remote*/nonusers

Subjects with current** SSRI use: HR 1.24; (95% CI 1.07 to 1.44)
Subjects with current tricyclic antidepressant use: HR 1.34; (95% CI 1.10 to 1.62)
Subjects with current use of other antidepressants: HR 1.43; (95% CI 1.21 to 1.69)

The risk of ischemic stroke in current* SSRI users was significantly higher compared with remote/nonusers; HR=1,55 (95% CI 1,00 to 2,39), while there was no significant risk of ischemic stroke in current users of TCAs or other antidepressants
The risk of hemorrhagic stroke in current users of an SSRI, TCA, or other antidepressant was not significantly different compared with that of remote/nonusers.

*remote use of antidepressant: antidepressant ended 91 or more than 91 days before the cerebrovascular event
**current use of antidepressant: antidepressant ended 30 days or less than 30 days before the cerebrovascular event.
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS:see results
QUALITY RATING:Good
STUDY:Authors: Cipriani A. et al.189
Year: 2006
Country: Multinational
FUNDING:No external funding- authors associated with Italian, Japanese and English universities
DESIGN:Study design: Systematic review and meta-analysis
Number of patients: 14391
AIMS OF REVIEW:To systematically review the efficacy and tolerability of fluoxetine, the most widely studied of newer antidepressants, in comparison with all other antidepressants in the acute treatment of depression in patients aged more than 18 years.
STUDIES INCLUDED IN REVIEW131 RCTs
TIME PERIOD COVERED:1966 to 2004
CHARACTERISTICS OF INCLUDED STUDIES:Published randomized trials, blind or open
CHARACTERISTICS OF INCLUDED POPULATIONS:Depressed patients 18 years or older
CHARACTERISTICS OF INTERVENTIONS:Fluoxetine in comparison with all other antidepressants in the acute treatment of depression.
MAIN RESULTS:Meta-analysis of Response Fluoxetine vs.
ADVERSE EVENTS:Meta-analysis of tolerability via all withdrawals Fluoxetine vs.
COMPREHENSIVE LITERATURE SEARCH STRATEGY:Cochrane Collaboration Depression, Anxiety and Neurosis Controlled Trials Register and the Cochrane Central Register of Controlled Trials up to March 2004; MEDLINE (1966–2004) and EMBASE (1974–2004)
STANDARD METHOD OF APPRAISAL OF STUDIES:Yes- Cochrane Collaboration Handbook
QUALITY RATING:Good
STUDY:Authors: Clayton A. et al. 21
Year: 2006
Country: USA
FUNDING:GlaxoSmithKline
DESIGN:Study design: 2 pooled RCTs
Setting: Multicenter
Sample size: 785 ITT
INTERVENTION:
Drug:Bupropion XLEscitalopramPlacebo
Dose:300–450 mg10–20 mgNA
Duration:8 weeks8 weeks8 weeks
Sample size:276281273
INCLUSION:Men and women > 18 years old, MDD; HAMD17 > 19,; current episode duration 12 weeks to 2 years; sexually active.
EXCLUSION:Other sexual disorders; past or present anorexia nervosa, bulimia, seizure disorder, or brain injury; diagnosis of panic disorder, OCD, PTSD or acute stress disorder within 12 months: bipolar I or II, schizophrenia or other psychotic disorders; attempted suicide within 6 months; any drug that may effect sexual functioning.
OTHER MEDICATIONS/ INTERVENTIONS:Zolpidem, zaleplon and and non-prescription sleep aids were allowed in 1st 10 days only.
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: Bupropion XL 37 Escitalopram 37 Placebo 36
Gender (female %): Bupropion XL 58 Escitalopram 57 Placebo 60
Ethnicity: White Bupropion XL 70% Escitalopram 68% Placebo 70%
Black Bupropion XL 20% Escitalopram 19% Placebo 17%
Other population characteristics: NR
OUTCOME ASSESSMENT:Primary Outcome Measures: % patients w/orgasm dysfunction at week 8
Secondary Outcome Measures: CSFQ, HAMD17, CGI-S and CGI-I and HAD
Timing of assessments: Baseline, weeks 1,2,3,4,6 and 8
RESULTS:
ANALYSIS:ITT: Yes
Post randomization exclusions: 45
Loss to follow-up differential high: No
ATTRITION:Bupropion XLEscitalopramPlacebo
Loss to follow-up:68 (25%)71 (25%)66 (24%)
Withdrawals due to adverse events:6%4%5%
Withdrawals due to lack of efficacy:NRNRNR
ADVERSE EVENTS:Bupropion XL vs. Escitalopram vs. Placebo %
QUALITY RATING:Fair
STUDY:Authors: Clayton AH, et al.190
Year: 2002
Country: US
FUNDING:Glaxo Wellcome Inc.
DESIGN:Study design: Cross sectional survey
Setting: Multi-center
Sample size: 6297
INTERVENTION:
Drug:Second generation antidepressants
Dose:Variable
Duration:Variable
INCLUSION:≥ 18 years of age; receiving antidepressant monotherapy for depression; sexually active; using one of the newer antidepressants: buproprion IR, buproprion SR, citalopram, fluoxetine, mirtazapine, nefazodone, paroxetine, sertraline, venlafaxine, venlafaxine XR
EXCLUSION:Taking an antidepressant for an illness other than depression
OTHER MEDICATIONS/ INTERVENTIONS:None
POPULATION CHARACTERISTICS:Groups similar at baseline: N/A
Mean age: Overall clinical population: 42.7; target population: 32.0 (target population consisted of patients free of other probable causes of sexual dysfunction (e.g., age, comorbid illness)
Gender (% female): overall clinical population: 28%; target population: 22.8%
Ethnicity: overall clinical population: white: 93.5%, black: 2.7%, Asian: 0.5%, Hispanic: 2.7%, other: 0.6%; target population: white: 93.1%, black: 2%, Asian: 0.6%, Hispanic: 3.7%, other: 0.5%
Other population characteristics: Not reported
OUTCOME ASSESSMENT:Measures: Changes in sexual functioning questionnaire
Timing of assessments: Completed at one visit
RESULTS:In the overall clinical population: In the target population:
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:N/A
QUALITY RATING:N/A
STUDY:Authors: Coleman CC, et al.22
Year: 1999
Country: US
FUNDING:Glaxo Wellcome
DESIGN:Study design: RCT
Setting: Multi-center (9 centers)
Sample size: 364
INTERVENTION:
Drug:SertralineBuproprionPlacebo
Dose:50–200 mg/d150–400 mg/dN/A
Duration:8 weeks8 weeks8 weeks
INCLUSION:DSM-IV criteria for major depression; minimum score of 18 on the first 21 items of the 31 item HAM-D; 18 years of age or older; be in a stable relationship, have normal sexual functioning, and sexual activity at least once every 2 weeks; currently experiencing recurrent major episode of duration 2–24 months
EXCLUSION:Predisposition to seizure or taking med that lowers seizure threshold; anorexia or bulimia; pregnancy; alcohol or substance abuse; eating disorder; suicidal tendencies; prior treatment with buproprion or sertraline; used any psychoactive drug within 1 week of study (2 weeks for MAOI or 4 weeks for fluoxetine)
OTHER MEDICATIONS/ INTERVENTIONS:Chloral hydrate for sleep (first 2 weeks only)
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: Sertraline: 38.3, buproprion: 38.1, placebo: 38.5
Gender (% female): 59%; sertraline: 54%, buproprion: 56%, placebo: 59%
Ethnicity: Sertraline: white: 92%, black: 8%, other: < 1%; buproprion: white: 87%, black: 11%, other: 2%; placebo: white: 88%, black: 9%, other: 3%
Other population characteristics: No significant differences at diagnosis
OUTCOME ASSESSMENT:Measures: 31 item HAM-D, HAM-A, CGI-S, CGI-I, sexual functioning by investigator questions: sexual desire disorder, sexual arousal disorder, orgasm dysfunction, premature ejaculation, patient rated overall sexual function
Timing of assessments: Baseline, weeks 1, 2, 3, 4, 6, 8
RESULTS:
  • Mean HAM-D scores in the buproprion but not the sertraline group were statistically better than placebo (by day 28 p < 0.05)
  • There was no significant difference between the buproprion and sertraline groups
  • CGI-I and CGI-S for buproprion significantly better than placebo but not better than sertraline
  • Sertraline not statistically better than placebo
  • No differences in HAM-A; significantly fewer buproprion patients had sexual desire disorder than sertraline patients (p < 0.05)
  • There was no significant difference between either active treatment group and placebo
  • Orgasm dysfunction occurred significantly more in sertraline patients compared with placebo or buproprion patients (p < 0.05)
  • Diagnosed with at least one sexual dysfunction: sertraline: 39%, buproprion: 13%, placebo: 17%
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes
ATTRITION:Loss to follow-up: 30%; sertraline: 36%, buproprion sr: 22%, placebo: 32%
Withdrawals due to adverse events: 18:5%; sertraline: 8%, buproprion: 6%, placebo: 2%
Loss to follow-up differential high: No
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Coleman CC, et al.23
Year: 2001
Country: US
FUNDING:Glaxo Wellcome
DESIGN:Study design: RCT
Setting: Multi-center (15 centers)
Sample size: 456
INTERVENTION:
Drug:BuproprionFluoxetinePlacebo
Dose:150–400 mg/d150–400 mg/dN/A
Duration:8 weeks8 weeks8 weeks
INCLUSION:DSM-IV criteria for major depression; minimum score of 20 on the 21 item HAM-D; ≥18 years of age; have sexual activity at least once every 2 weeks; currently experiencing episode lasting 2–24 months
EXCLUSION:Predisposition to seizure; pregnancy; alcohol or substance abuse; eating disorder; suicidal; treatment with buproprion or fluoxetine in the past year; used any psychoactive drug within 1 week of study; non-responders to antidepressant treatment; anorexia or bulimia
OTHER MEDICATIONS/ INTERVENTIONS:Not reported
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: Fluoxetine: 37.1, buproprion sr: 36.6, placebo: 36.7
Gender: (% female) Fluoxetine: 66%, buproprion: 63%, placebo: 61%
Ethnicity: Fuoxetine: white 82%, black 11%, other 7%; buproprion: white 83%, black 11%, other 5%; placebo: white 82%, black 14%, other 4%
Other population characteristics: At baseline more patients in the fluoxetine and buproprion goups than the placebo group had sexual desire disorder
OUTCOME ASSESSMENT:Measures: 21 item HAM-D, sexual function assessment, substance-induced arousal disorder and orgasm dysfunction. Assessed: orgasm dysfunction, sexual desire disorder, sexual arousal disorder, overall patient sexual functioning (1–6 scale)
Timing of assessments: Baseline, weeks 1, 2, 3, 4, 5, 6, 7, 8
RESULTS:
  • Mean HAM-D scores were not statistically different between the three groups (in ITT analysis)
  • No difference in responders (≥ 50 decrease in HAM-D), remitters (HAMD < 8)
  • More buproprion remitters (47%) compared to placebo (32%).
  • Orgasm dysfunction occurred significantly more in fluoxetine patients compared with placebo or buproprion patients (p < 0.001)
  • At endpoint more fluoxetine treated patients had sexual desire disorder than buproprion-treated patients (p < 0.05).
  • More fluoxetine-treated patients dissatisfied with sexual function beginning at week 1 (p < 0.05)
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes
ATTRITION:Loss to follow-up: 34%
Withdrawals due to adverse events: fluoxetine: 4%, buproprion: 9%, placebo: 3%
Loss to follow-up differential high: No
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Coogan PF, et al.191
Year: 2005
Country: US
FUNDING:NR
DESIGN:Study design: Case-control
Setting: 3 centers
Sample size: 4996
INTERVENTION:CasesControls
Drug:SSRIsNone
Dose:VariousN/A
Duration:N/AN/A
Sample size:21382858
INCLUSION:Cases: women with a first occurrence of primary invasive breast cancer diagnosed within the last year and no concurrent or previous cancer other than nonmelanoma skin cancer
Controls: women admitted for nonmalignant diagnoses, unrelated to the use of SSRIs and no history of cancer other than nonmelnoma skin cancer
EXCLUSION:N/A
OTHER MEDICATIONS/ INTERVENTIONS:N/A
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Range of age: 24–73
Gender (% female): 100%
Ethnicity: NR
OUTCOME ASSESSMENT:Primary Outcome Measures: Increased risk of breast cancer due to use of SSRIs

Risk factors other than SSRI use that were taken into account include alcohol consumption, religion, family history of breast cancer, center, age and race

Secondary Outcome Measures:

Timing of Assessments:
RESULTS:
  • Regular use of SSRIs was not associated with breast cancer risk after adjustment for other risk factors OR 1.1 95% 0.8, 1.7
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Withdrawals due to lack of efficacy: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Cornelius et al.192
Year: 2009
Country: USA
FUNDING:National Institute on Alcohol Abuse and Alcoholism, National Institute of Drug Abuse
DESIGN:Study design: RCT
Setting: Outpatient
Sample size: 50
INTERVENTION:
Drug:FluoxetinePlacebo
Dose:20 mg20 mg
Duration:12 weeks12 weeks
Sample size:2426
INCLUSION:15 and 20 years of age; DSM-IV confirmed diagnoses of current alcohol use disorder (AUD) and of current MDD (Schedule for Affective Disorders and Schizophrenia-Present and Lifetime Version (K-SADS-PL) used for MDD diagnosis; DSM-IV diagnosis of alcohol use disorder (alcohol abuse or dependence) was confirmed using the Substance Use Disorders Section of the Structured Clinical Interview (SCID).
Minimum levels of drinking for study inclusion were defined as drinking at least 10 drinks over the month prior to baseline assessment, as demonstrated on the Timeline Follow-back scale. HAM-D-27 score ≥15 at baseline assessment.
EXCLUSION:Bipolar disorder, schizoaffective disorder, or schizophrenia; hyper- or hypothyroidism, significant cardiac, neurological, or renal impairment, and significant liver disease; antipsychotic or antidepressant medication in the month prior to enrollment; any substance abuse or dependence other than nicotine dependence or cannabis abuse or dependence; history of intravenous drug use; pregnancy, inability or unwillingness to use contraceptive methods, and an inability to read or understand study forms.
OTHER MEDICATIONS/ INTERVENTIONS:Both groups: 9 sessions of manual-based intensive therapy, which consisted of Cognitive Behavioral Therapy (CBT) and Motivation Enhancement Therapy (MET).
POPULATION CHARACTERISTICS:Groups similar at baseline: Placebo significantly more depressed at baseline
Mean age: NR
Gender (female %): 50.0% fluoxetine, 61.5% placebo
Ethnicity (White %): 83.3% fluoxetine, 88.5% placebo
Other characteristics: Beck Depression Inventory (BDI) [mean score, and (SD)]: fluoxetine 17.28 (8.87) vs. placebo 22.12 (7.50), P < 0.041; Hamilton Rating Scale for Depression (HAM-D-27): fluoxetine 16.88(7.09) vs. placebo 22.88 (8.79), P < 0.011
OUTCOME ASSESSMENT:Primary Outcome Measures: Depressive symptoms: HAM-D-27 & BDI; drinking behavior (TLFB): drinks per day, drinks per occasion, days of alcohol use per week, heavy drinking days per week
Secondary Outcome Measures: NR
Timing of assessments: Baseline, weeks 1, 2, 3, 4, 6, 8, 10, and 12
RESULTS:No significant differences between fluoxetine and placebo in depressive symptoms or drinking behavior between the groups, with participants in both arms showing improvements for depressive symptoms and level of drinking.
Depressive symptoms: [mean score, (SD)]: BDI fluoxetine 6.79 (7.49) vs. placebo 10.46 (10.80), P = 0.173; HAM-D-27: fluoxetine 4.54 (7.06) vs. placebo 8.31 (8.97), P = 0.107.

Number of days of heavy alcohol use was significantly associated with lack of remission of BDI depression scores (BDI scores < 8) both at midpoint and end of study.
ANALYSIS:ITT: Yes
Post randomization exclusions: None
ATTRITION:Overall Attrition: 3/50; 6%
Withdrawals due to adverse events: 0
Withdrawals due to lack of efficacy: 3/50; 6% (all placebo)
Differential Attrition: 12% vs. 0%
ADVERSE EVENTS:No severe adverse events. Only mild and rare side effects occurred (no data reported).
QUALITY RATING:Fair
STUDY:Authors: Croft H, et al.27
Year: 1999
Country: US
FUNDING:Glaxo Wellcome
DESIGN:Study design: RCT (active and placebo control)
Setting: Multi-center (8 centers)
Sample size: 360
INTERVENTION:
Drug:SertralineBuproprionPlacebo
Dose:50–200 mg/d150–400 mg/dN/A
Duration:8 weeks8 weeks8 weeks
INCLUSION:DSM-IV criteria for major depression; minimum score of 18 on the first 21 items of the 31 item HAM-D; ≥ 18 years of age; in a stable relationship; have normal sexual functioning and sexual activity at least once every 2 weeks; current depressive episode of 8 weeks to 24 months
EXCLUSION:Predisposition to seizure; pregnancy; alcohol or substance abuse; eating disorder; suicidal tendencies; prior treatment with buproprion or sertraline; used any psychoactive drug within 1 week of study
OTHER MEDICATIONS/ INTERVENTIONS:Not reported
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: Sertraline: 36.0, buproprion: 35.9, placebo: 37.4
Gender (% female): Sertraline: 50%, buproprion: 51%, placebo: 50%
Ethnicity: Sertraline: white: 87%, black: 8%, other: 4%; buproprion: white: 86%, black: 9%, other: 5%; placebo: white: 88%, black: 8%, other: 3%
Other population characteristics: Not reported
OUTCOME ASSESSMENT:Measures: 31 item HAM-D, HAM-A, CGI-S, CGI-I, sexual function assessment by investigator interview-sexual desire disorder, sexual arousal disorder, orgasmic dysfunction, premature ejaculation (men only), overall patient satisfaction with sexual functioning, vital signs
Timing of assessments: Baseline, weeks 1, 2, 3, 4, 6, 8
RESULTS:
  • Mean HAM-D scores in both the buproprion and sertraline group were statistically better than placebo (p < 0.05)
  • No significant difference in HAM-D scores between the buproprion and sertraline groups
  • CGI-S and CGI-I improvement compared to placebo but no differences between drugs at any week
  • No difference in changes of HAM-A scores for any group
  • By day 42 significantly fewer buproprion sr-treated patients had sexual desire disorder than sertraline- or placebo-treated patients (p < 0.05)
  • At day 56 both buproprion and sertraline groups had higher sexual arousal disorder (p < 0.05) than placebo
  • Orgasmic dysfunction occurred significantly more in sertraline group compared with placebo or buproprion groups (p < 0.001)
  • At day 56 no difference in overall satisfaction with sexual function between treatment groups
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes
ATTRITION:Loss to follow-up: 32%
Withdrawals due to adverse events: sertraline: 3%, buproprion sr: 3%, placebo: 7%
Loss to follow-up differential high: Yes
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: de Abajo et al.193
Year: 2008
Country: United Kingdom
FUNDING:unrestricted research grant from Astra Zeneca PLC for the validation of cases
DESIGN:Study design: Nested case control study
Setting: UKPDS - The Health Improvement Network database in the United Kingdom
Sample size: 1321 cases, 10,000 controls
INTERVENTION: = exposureCases (having an upper GI tract complication when no exclusion criteria were found up to 2 months after the computer date of case detection)Controls
Drug:SSRIs & SNRIs (Sertraline, Fluoxetine, Fluvoxamine, Paroxetine, Citalopram, Escitalopram, Venlafaxine, Duloxetine)SSRIs & SNRIs (Sertraline, Fluoxetine, Fluvoxamine, Paroxetine, Citalopram, Escitalopram, Venlafaxine, Duloxetine)
Dose:Low vs. medium to high usersLow vs. medium to high users
Duration:NANA
Sample size:1,32110,000
INCLUSION:Persons aged 40 to 84 years who have been seen for at least 2 years by a general practitioner and with at least 1 year elapsed since the first recorded prescription. Cases were having an upper gastrointestinal (GI) tract complication. Control patients were randomly selected from the source population using density-based sampling method matched for age, sex and calendar year of the index date.
EXCLUSION:History of cancer, liver disease, coagulopathy, Mallory-Weiss syndrome, esophageal varices, or alcohol-related disorders were excluded. In addition, persons aged 70 years or older at the start date with a follow-up of longer than 1 year and with no recording of data or with only 1 medical visit.
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: No – more current smokers (20.5% vs. 15.3%) and current users (19.5% vs. 11.2%) as well as past users (7.5% vs. 3.5%) of acid suppressing agents in cases and more people suffering from antecedents of GI tract disorders (43.4% vs. 23.9%) in cases.
Mean age: cases: 40–59 years (24.5%); 60–69 (22.9%); 70–79 (36.5%); 80–84 (16.4%) controls: 40–59 years (27.6%); 60–69 (21.5%); 70–79 (34.5%); 80–84 (16.5%)
Gender (female %): 41.6% cases vs. 43.3% controls
Ethnicity (Caucasian %): NR
Other population characteristics: NA
OUTCOME ASSESSMENT:Primary Outcome Measures: upper GI Tract Bleeding
Secondary Outcome Measures: upper GI Tract Perforation
Timing of assessments: observation period (January 1, 2000 to December 31, 2005)
RESULTS:The percentages of current users of SSRIs was higher in cases than in controls (5.3% vs. 3.0%). Suffering from GI tract bleeding is associated with taking SSRIs, yielding adjusted ORs of 1.6 (95% CI, 1.2 to 2.1).
The percentages of current users of SNRIs was higher in cases than in controls (1.1% vs. 0.3%), yielding adjusted ORs of 2.9 (95% CI, 1.5 to 5.6).
There was a higher chance of cases taking Sertraline OR: 2.3 (95% CI, 1.0 to 5.1), Citalopram or Escitalopram OR: 2.0 (95% CI, 1.2 to 3.2), Venlafaxine OR: 2.9 (95% CI, 1.5 to 5.7).
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS: see results
QUALITY RATING:Fair
STUDY:Authors: Didham RC, et al.194
Year: 2005
Country: New Zealand
FUNDING:The Royal NZ College of General Practitioners Research Unit which receives funding from the NZ government
DESIGN:Study design: Retrospective cohort and nested case control study
Setting: General practice
Sample size: 57,361
INTERVENTION:
Drug:SSRIs and other ADS
Dose:Varied
Duration:120 days
Cases:Suicides: 26
Self-harms: 330
INCLUSION:Patients that received a prescription for an anti-depressant from 1996 to 2001
EXCLUSION:Patients under 10 years old; additional concurrent anti-depressants
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Median age: 46
Gender (% female): 68.1%
Ethnicity: NR
OUTCOME ASSESSMENT:Primary Outcome Measures: Suicides or self-harm within 120 days of a prescription

Timing of assessments: N/A
RESULTS:
  • No significant increase in suicides for SSRIs as a group: OR 1.28; 95% CI 0.38–4.35
  • No significant difference in suicides between drugs
    Fluoxetine: 0.80 (0.22–2.89)
    Paroxetine: 2.25 (0.47–10.72)
  • Self-harm SSRIs vs. TCAs incidence rate 2.57 95% CI 2.03–3.28
  • Increased risk of self-harm for SSRIs as a group OR 1.66 95% CI 1.23–2.23
  • No significant differences in self-harm between drugs
    Fluoxetine; 1.30 (0.96–1.75)
    Paroxetine 1.21 (0.84–1.72)
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Withdrawals due to lack of efficacy: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Dunner et al.195
Year: 1998
Country: US
FUNDING:Glaxo Wellcome Inc., Research Triangle Park, NC
DESIGN:Study design: Observational prospective
Setting: Multi-center (105 sites)
Sample size: 3100
INTERVENTION:Bupropion
Drug:
Dose:100–300 mg/d
Duration:8 weeks
Sample size:3100
INCLUSION:Male or female patients at least 18 years of age; met DSM-III-R criteria for MDD, dysthymia, bipolar I or II)
EXCLUSION:Previous treatment with bupropion; patients with a history of bulimia or anorexia or with a known predisposition to seizures; pregnant; lactating; suicidal
OTHER MEDICATIONS/ INTERVENTIONS:Benzodiazepines
POPULATION CHARACTERISTICS:Groups similar at baseline: N/A
Mean age: 42
Gender (% female): 62.4
Ethnicity: white: 89.5%, black: 7%, other: 3.5%
Other population characteristics: NR
OUTCOME ASSESSMENT:Primary Outcome Measures: Number of seizures; seizure rate

Secondary Outcome Measures: N/A

Timing of assessments: Biweekly during the study
RESULTS:
  • During the 8 week acute phase of the trial, 2 patients (0.06% -- Upper 1-sided CL of 0.14%) experienced seizures out of 3094 patients.
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Overall
Loss to follow-up:34%
Withdrawals due to adverse events:NR
Withdrawals due to lack of efficacy:NR
Loss to follow-up differential high:N/A
ADVERSE EVENTS:
  • 54 serious adverse events (other than seizure) occurred during the study. Suicide attempt or overdose: 9 patients; accidental injury: 4 patients; myocardial function: 3 patients
QUALITY RATING:Fair
STUDY:Authors: Ekselius, et al.196
Year: 2001
Country: Sweden
FUNDING:Swedish Medical Research Council and Pfizer AB
DESIGN:Study design: Subgroup analysis of RCT
Setting: Multi-center
Sample size: 400
INTERVENTION:
Drug:SertralineCitalopram
Dose:50–150 mg/d20–60 mg/d
Duration:24 weeks24 weeks
INCLUSION:DSM-III-R criteria for major depression; MADRS score ≥ 21
EXCLUSION:Pregnancy; alcohol or substance abuse; suicidal tendencies; significant physical illness; bipolar disorder; known intolerance or allergic reactions to SSRIs; severe depression or psychotic dimension; previous adequate treatment with citalopram or sertraline; lithium within past month
OTHER MEDICATIONS/ INTERVENTIONS:Hypnotics for insomnia or daytime anxiolytics
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Gender (% female): Sertraline: 72%, citalopram: 71%
Ethnicity: Not reported
Mean age: Sertraline: 47.3, citalopram: 48.1
Other population characteristics: No significant population differences
OUTCOME ASSESSMENT:Measures: MADRS, CGI-S, CGI-I, sexual function assessed by five items in the Utvalg for Kliniske Undersogelser Side Effect Scale (UKU-SES); increased or decreased sexual desire, erectile dysfunction, ejaculatory dysfunction, orgasmic dysfunction
Timing of assessments: Not reported
RESULTS:
  • No statistically significant differences between sertraline and citalopram in the magnitude or frequency of adverse sexual side effects
  • For both groups sexual desire and mean total score of UKU significantly improved in women; sexual desire improved in men, but not mean score of UKU.
  • In female patients reporting no sexual dysfunction at baseline, 11.8% reported decreased sexual desire and 14.3% reported orgasmic dysfunction
  • In male patients reporting no sexual dysfunction at baseline, 16.7% reported decreased sexual desire, 18.9% reported orgasmic dysfunction, 25% experienced ejaculatory dysfunction
ANALYSIS:ITT: Not reported
Post randomization exclusions: Not reported
ATTRITION:Loss to follow-up: 23%; sertraline: not reported, citalopram: not reported
Withdrawals due to adverse events: 11%; sertraline: not reported, citalopram: not reported
Loss to follow-up differential high: Not reported
ADVERSE EVENTS:Not reported
QUALITY RATING:Fair
STUDY:Authors: Fava M, et al.36
Year: 2002
Country: US
FUNDING:Eli Lilly Research
DESIGN:Study design: RCT
Setting: Multi-center
Sample size: 284
INTERVENTION:
Drug:FluoxetineSertralineParoxetine
Dose:20–60 mg/day50–200 mg/day20–60 mg/day
Duration:10–16 weeks10–16 weeks10–16 weeks
INCLUSION:≥ 18 years of age; DSM-V criteria for major depression; DSM-IV for atypical MDD; HAM-D-17 ≥16; episode ≥1 month
EXCLUSION:Pregnancy or lactation, lack of adequate contraception; history of psychotic disorders, bipolar disorder; alcohol or substance abuse; existing suicidal risk; previously failed to respond to antidepressant therapies; clinically relevant progressive disease; hypersensitivity to study medication; serious comorbid illness not stabilized; anxiolytic or psychotropic within 7 days; MAOI within 2 weeks
OTHER MEDICATIONS/ INTERVENTIONS:Thyroid medications, chloral hydrate
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: Fluoxetine: 42.1, sertraline: 44.0, paroxetine: 42.5
Gender (female%): Fluoxetine: 63.0, sertraline: 57.3, paroxetine: 58.3
Ethnicity: Not reported
Other population characteristics: Not reported
OUTCOME ASSESSMENT:Measures: HAM-D-17, CGI-S, HAM-D sleep disturbance
Timing of assessments: Not reported
RESULTS: Subgroup analysis (Fava 2000): Anxious depression
ANALYSIS:ITT: Yes
Post randomization exclusions: Not reported
ATTRITION:Loss to follow-up: 27.1%; fluoxetine: 26.1%, sertraline: 27.1%, paroxetine: 28.1%
Withdrawals due to adverse events: Fuoxetine: 8.7%, sertraline: 6.3%, paroxetine: 11.5%
Loss to follow-up differential high: No
ADVERSE EVENTS:
  • Pairwise comparisons indicated that the paroxetine-treated patients reported more constipation than the fluoxetine-treated patients; the fluoxetine-treated patients reported more twitching and cough increase than the sertraline-treated patients
  • Most common adverse events: Fluoxetine: headache (25%); sertraline: headache (28.1%), diarrhea (26.0%), insomnia (26%), nausea (20.8%); paroxetine: nausea (25.0%), headache (21.9%), insomnia (20.8%), abnormal ejaculation (20.8%)
  • There was a significant increase in weight for the paroxetine group; fluoxetine treated patients showed a significant decrease in weight and the sertraline group a non-significant decrease in weight from baseline to endpoint
Subgroup analysis (Fava 1999)
  • Adverse events were similar among treatments; only flu-like syndrome was significantly higher in the sertraline treated group overall (p = 0.021)
QUALITY RATING:Fair
STUDY:Authors: Fergusson D, et al.197
Year: 2005
Country: Canada
FUNDING:Canadian Institutes of Health Research
DESIGN:Study design: Meta-analysis
Number of patients: 36,445
AIMS OF REVIEW:To establish if an association exists between SSRI use and suicide attempts.
STUDIES INCLUDED IN META- ANALYSIS345 trials included in analysis
TIME PERIOD COVERED:1967 – June 2003
CHARACTERISTICS OF INCLUDED STUDIES:RCTs comparing an SSRI with either placebo or an active non-SSRI control
CHARACTERISTICS OF INCLUDED POPULATIONS:All patients included in trials comparing SSRIs to either placebo or non-SSRI control; no age, gender, or diagnosis restrictions
CHARACTERISTICS OF INTERVENTIONS:Patients randomized to either an SSRI, placebo, or non-SSRI control
MAIN RESULTS:
  • A significant increase in the odds of suicide attempts was found in patients receiving SSRIs compared to patients receiving placebo (OR: 2.28; CI: 1.144 to 4.55; p = 0.02)
  • No significant difference found in the odds of suicide attempts between patients receiving SSRIs and patients receiving TCAs (OR: 0.88 (CI: 0.54 to 1.42)
ADVERSE EVENTS:
  • No other adverse events reported.
COMPREHENSIVE LITERATURE SEARCH STRATEGY:Yes
STANDARD METHOD OF APPRAISAL OF STUDIES:Yes
QUALITY RATING:Good
STUDY:Authors: Gartlehner et al 198
Year: 2008
Country: Austria / USA
FUNDING:AHRQ
DESIGN:Study design: Systematic Review and Metaanalysis
Number of patients: > 757000
AIMS OF REVIEW:To review systematically the comparative harms of second generation antidepressants for the treatment of MDD
STUDIES INCLUDED IN REVIEW83 head to head RCTs (81 double blinded, two open label), 21 observational studies
TIME PERIOD COVERED:1980 - April 2007
CHARACTERISTICS OF INCLUDED STUDIES:Experimental and observational head-to-head studies with a minimum duration of 6 weeks
CHARACTERISTICS OF INCLUDED POPULATIONS:Adult patients with MDD
CHARACTERISTICS OF INTERVENTIONS:Bupropion, citalopram, duloxetine, escitalopram, fluoxetine, fluvoxamine, mirtazapine, nefazodone, paroxetine,, sertraline, trazodone, venlafaxine
MAIN RESULTS:Serious adverse events:
ADVERSE EVENTS:See above
COMPREHENSIVE LITERATURE SEARCH STRATEGY:Yes
STANDARD METHOD OF APPRAISAL OF STUDIES:Yes
QUALITY RATING:Good
STUDY:Authors: Gibbons RD et al.199
Year: 2007
Country: USA
FUNDING:NIMH
DESIGN:Study design: Observational – retrospective cohort
Setting: VA hospitals database
Sample size: 226,866
INTERVENTION:
Drug:No anti-depressantSSRI monotherapyNon-SSRI monotherapy
Dose:NAVariousVarious
Duration:6 months6 months6 months
Sample size:59,43282,82827,548
(bupropion, mirtazapine, nefazodone, and Venlafaxine)
INCLUSION:Depressive disorders or unipolar mood disorders in 2003 or 2004, had at least 6 months of follow-up, and had no history of these disorders or antidepressant treatment from 2000 to 2002
EXCLUSION:NA
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline:
Mean age: No anti-depressant 57.6 SSRI 60.3 Non-SSRI 55.6
Gender (female %): No anti-depressant 8.4 SSRI 7.8 Non-SSRI 7.3
Ethnicity: % black No anti-depressant 8.3 SSRI 5.3 Non-SSRI 6.8
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Suicide attempts
Secondary Outcome Measures:
Timing of assessments: 6 months
RESULTS:Suicide attempt rates were lower among patients who were treated with antidepressants than among those who were not, with a statistically significant odds ratio for SSRIs and tricyclics. For SSRIs versus no antidepressant, this effect was significant in all adult age groups.

Age group no anti depressant vs SSRI monotherapy Odds ratio (95% CI) p value
18–25 0.35 (0.14–0.85) p = 0.021
0.44 (0.29–0.65) p < 0.0001
46–65 0.42 (0.30–0.59) p < 0.0001
>65 0.38 (0.16–0.91) p = 0.036

Treatment compared to no treatment, likelihood of suicide attempt
No antidepressant Attempts = 199 Rate per 100,000 =335
SSRI monotherapy Attempts = 102 Rate per 100,000= 123 OR = 0.37 95% CI 0.29–0.47 P <0.0001
Non-SSRI monotherapy Attempts = 76 Rate per 100,00 = 276 OR = 0.83 95% CI 0.64–1.08 P = 0.16
ANALYSIS:ITT: NA
Post randomization exclusions: NA
Loss to follow-up: NA
ATTRITION:NA
Withdrawals due to adverse events:
Withdrawals due to lack of efficacy:
Loss to follow-up differential high:
ADVERSE EVENTS:
  • See results
QUALITY RATING:Fair
STUDY:Authors: Greist J, et al.200
Year: 2004
Country: US
FUNDING:Eli Lilly
DESIGN:Study design: Pooled analysis
Number of patients: 2,345
AIMS OF REVIEW:To assess the incidence, severity and onset of nausea among MDD patients treated with duloxetine
STUDIES INCLUDED IN META-ANALYSISDetke et al. 2002; Detke et al. 2002; Goldstein et al 2002; Goldstein et al. 2004; 4 unpublished studies submitted for FDA approval of duloxetine
TIME PERIOD COVERED:Not reported
CHARACTERISTICS OF INCLUDED STUDIES:Double blinded, placebo or active controlled trials of duloxetine
CHARACTERISTICS OF INCLUDED POPULATIONS:Adult outpatients with MDD
CHARACTERISTICS OF INTERVENTIONS:Duloxetine vs. placebo (8 studies); duloxetine vs. paroxetine (4 studies); duloxetine vs. fluoxetine (2 studies)
MAIN RESULTS:
ADVERSE EVENTS:N/A
COMPREHENSIVE LITERATURE SEARCH STRATEGY:No; analysis of published and unpublished trials
STANDARD METHOD OF APPRAISAL OF STUDIES:Not reported
QUALITY RATING:Fair
STUDY:Authors: Gunnell D, et al.201
Year: 2005
Country: UK
FUNDING:Not Reported
DESIGN:Study design: Meta-analysis
Number of patients: 40,826
AIMS OF REVIEW:To investigate whether SSRIs are associated with an increased risk of suicide related outcomes in adults.
STUDIES INCLUDED IN META-ANALYSISPublished and unpublished data submitted by pharmaceutical companies to the Medicine and Healthcare Products Regulatory Agency (MHRA) (2004)
342 placebo controlled trials included in report – citations not given in bibliography
TIME PERIOD COVERED:NR
CHARACTERISTICS OF INCLUDED STUDIES:Randomized, placebo controlled trials of SSRIs (citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, and sertraline) submitted by pharmaceutical companies
CHARACTERISTICS OF INCLUDED POPULATIONS:Adult patients with various indications included in trials comparing SSRIs to placebo.
CHARACTERISTICS OF INTERVENTIONS:Patients randomized to either SSRI or placebo.
MAIN RESULTS:
  • No significant difference was found between SSRI treatment and placebo treatment in the odds ratios for suicide (OR: 0.85 CI: 0.2 to 3.4), non-fatal self harm (OR: 1.57 CI: 0.99 to 2.55), or suicidal thought (OR: 0.77 CI: 0.37 to 1.55).
  • For non-fatal self-harm the NNT to harm is 759
ADVERSE EVENTS:
  • No other adverse events reported.
COMPREHENSIVE LITERATURE SEARCH STRATEGY:No (published and unpublished data submitted by pharmaceutical companies; review does not include studies from sources other than pharmaceutical companies)
STANDARD METHOD OF APPRAISAL OF STUDIES:Yes
QUALITY RATING:Good
STUDY:Authors: Hammad TA et al.202
Year: 2006
Country: USA
FUNDING:CDER, FDA
DESIGN:Study design: Meta-analysis
Number of patients: 4582
AIMS OF REVIEW:The objective of this article is to provide the detailed methods and results of the FDA’s exploration and analysis of the pediatric suicidality adverse event data and suicide item score data.
STUDIES INCLUDED IN REVIEW23 trials and 1 multicenter trial (TADS)
TIME PERIOD COVERED:NA - Most of the trials were conducted in the late 1990s, and trial durations ranged from 4 to 16 weeks.
CHARACTERISTICS OF INCLUDED STUDIES:23 placebo-controlled clinical trials conducted in 9 drug development programs of antidepressants in pediatric patients and in a placebo-controlled, multicenter trial funded by the National Institute of Mental Health
CHARACTERISTICS OF INCLUDED POPULATIONS:Children and adolescents with MDD (16 trials), obsessive-compulsive disorder (4 trials), generalized anxiety disorder (2 trials), social anxiety disorder (1 trial), and attention-deficit/hyperactivity disorder (1 trial).
CHARACTERISTICS OF INTERVENTIONS:Fluoxetine, sertraline hydrochloride, paroxetine, fluvoxamine maleate, citalopram hydrobromide, bupropion hydrochloride, venlafaxine hydrochloride (extended release), nefazodone hydrochloride, and mirtazapine.
MAIN RESULTS:
  • Overall Suicidal Behavior or Ideation Risk Ratio (95% CI) 1.95 (1.28 – 2.98)
ADVERSE EVENTS:MDD Trials RR (95% CI)All trials, all indications RR (95% CI)
Citalopram 1.37 (0.53–3.50)Citalopram 1.37 (0.53–3.50)
Fluvoxamine No MDD trialsFluvoxamine 5.52 (0.27–112.55)
Paroxetine 2.15 (0.71–6.52)Paroxetine 2.65 (1.00–7.02)
Fluoxetine 1.53 (0.74–3.16)Fluoxetine 1.52 (0.75–3.09)
Sertraline 2.16 (0.48–9.62)Sertraline 1.48 (0.42–5.24)
Venlafaxine ER 8.84 (1.12–69.51)Venlafaxine ER 4.97 (1.09–22.72)
Mirtazapine 1.58 (0.06–38.37)Mirtazapine 1.58 (0.06–38.37)
Nefazodone No eventsNefazodone No events
Bupropion No MDD trialsBupropion No events
COMPREHENSIVE LITERATURE SEARCH STRATEGY:No- request was from FDA to drug companies
STANDARD METHOD OF APPRAISAL OF STUDIES:NA - Patient level data
QUALITY RATING:Good
STUDY:Authors: Haffmans, et al.203
Year: 1996
Country: The Netherlands
FUNDING:Lundbeck
DESIGN:Study design: RCT
Setting: Multi-center
Sample size: 217
INTERVENTION:
Drug:CitalopramFluvoaxamine
Dose:20–40 mg/d100–200 mg/d
Duration:6 weeks6 weeks
INCLUSION:Ages 18–70 years; met DSM III-R criteria for major depression (single episode or recurrent) or bipolar disorder; score of ≥ 16 on HAM-D-17; reasonable knowledge of the Dutch language
EXCLUSION:MAOI or fluoxetine use within 3 weeks or other psychotropic drugs within 1 week (except for benzos); other primary psychiatric diagnosis (other than MDD); history of epilepsy, alcohol or drug abuse; pregnancy, lactation, or not using contraception; renal, hepatic, cardiovascular, neurological or somatic disorders and/or significant abnormal lab findings
OTHER MEDICATIONS/INTERVENTIONS:Selected benzodiazepines; oxazepam, lormetazepam, temazepam, lorazepam, or flurazepam, all non-psychotropic medications were allowed, domperidone for nausea/vomiting allowed
POPULATION CHARACTERISTICS:Groups similar at baseline: No
Mean age: Citalopram: 44.2, fluvoxamine: 40.2
Gender (% female): 58%; citalopram: 58%, fluvoxamine: 60%
Ethnicity: Not reported
Other population characteristics: Previous depressive disorder: citalopram: 43%; fluvoxamine: 54%; previous antidepressant therapy (within 3 weeks of starting trial): citalopram: 65%, fluvoxamine: 73%
OUTCOME ASSESSMENT:Measures: Primary: HAM-D-17; secondary: CGI, UKU side effect rating scale, Zung self-rating depression scale
Timing of assessments: Baseline, weeks 1, 2, 4, 6
RESULTS:
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes
ATTRITION:Loss to follow-up: 23%; citalopram: 19.4%, fluvoxamine: 26.6%
Withdrawals due to adverse events: Citalopram: 13.9%, fluvoxamine: 21.1%
Loss to follow-up differential high: No
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Isacsson G, et al.204
Year: 2005
Country: Sweden
FUNDING:The Soderstrom-Konigska Foundation and Karolinska Institute
DESIGN:Study design: Controlled database study
Setting:
Sample size: 41,279
INTERVENTION:
Drug:CasesControls
Dose:N/AN/A
Duration:9 year period9 year period
Sample size:14,85726,422
INCLUSION:Cases: suicide (as a Swedish citizen) investigated by the Department of Forensic Chemistry of the National Board of Forensic Medicine in Sweden where analysis detected therapeutic concentration of antidepressants in femoral blood; includes uncertain cases (overdose that may have been suicide)
Controls: investigated death during same time period which, after forensic investigation, was judged to be natural or accidental
EXCLUSION:N/A.
OTHER MEDICATIONS/INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: yes
Median age: cases: 49, controls: 55
Gender (female %): cases: 29%, controls: 27%
Ethnicity: 100%ll Swedish citizens (no further ethnicity reported)
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Detection of antidepressants in toxicological screening
Secondary Outcome Measures:
Timing of assessments: N/A
RESULTS:
  • 3,411 detections of antidepressants in suicides (cases) vs. 1,538 in controls
  • SSRIs underrepresented compared to other antidepressants (OR=0.83, 99% CI: 0.77–0.90)
  • SSRIs had lower OR (99% CI) than other antidepressants; citalopram: 0.76 (0.69–0.84), fluoxetine: 0.91 (0.60–1.38), fluvoxamine: 3.04 (1.15–8.04), paroxetine: 0.87 (0.60–1.28), sertraline: 1.05 (0.78–1.42)
  • Differences within SSRIs were insignificant with the exception of fluvoxamine
  • Other modern antidepressants (OR, 99%CI): mirtazapine: 1.67 (1.08–2.60), venlafaxine: 1.47 (0.99–2.18)
  • Excluding uncertain suicides from analysis changed Ors only marginally (data NR)
  • 52 suicides in people under 15 yrs of age but no SSRIs detected; venlafaxine detected in 1 case)
  • Among the 998 controls under 15 yrs of age, 4 were positive for antidepressants (3 for citalopram); SSRIs vs. non-SSRIs in cases and controls p=0.02
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Withdrawals due to lack of efficacy: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:N/A
QUALITY RATING:Fair
STUDY:Authors: Jick H, et al.205
Year: 2004
Country: UK
FUNDING:Boston Collaborative Drug Surveillance Program
DESIGN:Study design: Matched case-control; post-hoc database analysis
Setting: General practices in the UK using VAMP database (General Practice Research Database)
Sample size: 159,810 (555 cases, 2062 controls)
INTERVENTION:
Drug:Dothiepin, amitryptyline, fluoxetine, paroxetine
Dose:Not reported
Duration:Not reported
INCLUSION:Received a prescription for at least 1 antidepressant in the VAMP database during the 1993–1999 years; all patients who had a first-time recorded diagnosis of nonfatal suicidal ideation or attempted suicide at age 10–69 years during the 1993–1999 time period; had received at least 1 prescription for a study drug within 90 days before their index date
EXCLUSION:Received prescription for another antidepressant or more than one study drug prior to their index date; history of psychosis, panic disorders, phobias, obsessive-compulsive neurosis, manic-depressive disease, drug abuse, alcohol abuse, epilepsy, anorexia, bulimia, and attention-deficit disorder
OTHER MEDICATIONS/INTERVENTIONS:Not reported
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: not reported
Gender (% female): 65.4% female (cases only)
Ethnicity: Not reported
Other population characteristics: ~85% of cases had attempted suicide while 15% had suicidal ideation
OUTCOME ASSESSMENT:Measures: Frequency of first-time exposure to amitriptyline, fluoxetine, paroxetine and dothiepin of patients with a recorded diagnosis of first-time nonfatal suicidal behavior or suicide compared with matched patients who did not exhibit suicidal behavior
Timing of assessments: N/A
RESULTS:
  • Risk of suicidal behavior was similar among users of amitryptyline (RR: 0.83; 95% CI 0.61 – 1.13), fluoxetine (RR 1.16; 95% CI 0.90 – 1.50), and paroxetine (RR 1.29; 95% CI 0.97 – 1.70) compared to dotiepin
  • Suicide risk was increased in the first month after starting antidepressants, especially during the first 1 – 9 days (RR 4.07; 95% CI 2.89 – 5.74)
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:Not reported
QUALITY RATING:N/A
STUDY:Authors: Jick, et al.206
Year: 1995
Country: UK
FUNDING:Various pharmaceutical companies (Berlex, Boots, Burroughs Wellcome, Ciba-Geigy, Hoeschst, Hoffman-LaRoche, RW Johnson, Pfizer, Proctor and Gamble, Sanofi Winthrop
DESIGN:Study design: Cohort study with nested case-control analysis
Setting: General practices in the UK using VAMP database
Sample size: 172,598
INTERVENTION:
Drug:Drugs studies in this cohort: dothiepin, amitryptyline, climipramine, imipramine, flupenthixol, lofepramine, mianserin, fluoxetine, doxepin, trazodone, maprotiline, desipramine
Dose:Not reported
Duration:Not reported
INCLUSION:Received a prescription for 1 or more antidepressant in the VAMP database (General Practice Research Database); all patients who committed suicide identified in the cohort evaluation were included as cases
EXCLUSION:Not reported
OTHER MEDICATIONS/INTERVENTIONS:Not reported
POPULATION CHARACTERISTICS:Groups similar at baseline: Not reported
Mean age: Not reported
Gender: Not reported
Ethnicity: Not reported
Other population characteristics: Not reported
OUTCOME ASSESSMENT:Measures: Suicide completion rate, suicides/person time at risk, relative risks of suicide reported with dothiepin as reference group
Timing of assessments: N/A
RESULTS:
  • From cohort analysis: Suicide rate/10,000 person years: fluoxetine: 19.0, adjusted RR: 2.1 (95% CI 1.1–4.1) relative to dothiepin
  • From case control analysis: Adjusted RR 3.8 (95% CI 1.7–8.6), analysis restricted to those prescribed antidepressants for the first time and who had no history of suicidal behavior, adjusted RR: 2.1 (95% CI 0.6 – 7.9)
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: Not reported
Withdrawals due to adverse events: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:Not reported
QUALITY RATING:Fair
STUDY:Authors: Jick and Li 207
Year: 2008
Country: United Kingdom
FUNDING:NR
DESIGN:Study design: Nested case-control study
Setting: United Kingdom - based General Practice Research Database
Sample size: 3867
INTERVENTION:CasesControls
Drug:Tricyclic antidepressants, SSRIs and other antidepressants (details under “inclusion”)Tricyclic antidepressants, SSRIs and other antidepressants (details under “inclusion”)
Dose:variousvarious
Duration:1990–20051990–2005
Sample size:7823085
INCLUSION:All people in the database aged 70 years or younger who had filled at least one prescription for an antidepressant drug between 1990 and 2005 with a first time diagnosis of venous thromboembolism. To be considered as having a confirmed case of venous thromboembolism all subjects were required to have been either hospitalized or referred to a specialist and to have received anticoagulans. The patients had to have at least 1 year of information in the computer before the index date. (date of diagnosis)
EXCLUSION:Subjects with a history of trauma, surgery, or pregnancy within the 3 months before the index date were excluded from further study, as were subjects with a history of stroke, myocardial infarction or angina, cerebrovascular disease, epilepsy, renal failure, insulin-dependent diabetes mellitus, cancer, drug abuse, or alcohol abuse any time before the index date of case. Those with a history of anticoagulation therapy more than 60 days before the index date were also excluded
OTHER MEDICATIONS/INTERVENTIONS:Antipsychotic drugs, oral contraceptives, hormone replacement therapy
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes, except for the Body Mass Index, the current hormone replacement therapy use and the current oral contraceptive use
Mean age: NR
Gender (female %): 65,2%
Ethnicity (Caucasian %): NR
Other population characteristics: The current use of hormone replacement therapy and contraceptive use was higher in female case patients. More case patients than controls had the highest BMI (>=30)
More controls than cases had the lowest BMI. (<25)
OUTCOME ASSESSMENT:Primary Outcome Measures: Venous thromboembolism
Secondary Outcome Measures: None
Timing of assessments: NA
RESULTS:There was no overall effect of current antidepressant use on the risk of venous thromboembolism (OR 1.2, 95% CI 0.9–1.4)
Unadjusted ORs for current use of SSRIs compared with nonusers of any antidepressant (past use and nonusers combined): 0.9 (95% CI 0.6–1.2)
For current users of tricyclic antidepressants compared with nonusers, the unadjusted OR was 1.4 (95% CI 1.1–1.8)
For users of other antidepressants compared with nonusers the unadjusted OR was 1.0 (95% CI 0.5–2.0)
The unadjusted ORs of the effects for recent use were 1.2 (95% CI 0.6–2.8) for SSRI use and 1.3 (95% CI 0.7–2.5) for tricyclic antidepressants use compared with nonuse.
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS:See results
QUALITY RATING:Fair
STUDY:Authors: Johnston et al.208
Year: 1991
Country: US
FUNDING:Burroughs Wellcome Co., RTP, NC
DESIGN:Study design: Prospective observational
Setting: Multi-center (102 sites)
Sample size: 3341
INTERVENTION:Buproprion
Dose:225–450 mg/d
Duration:8 weeks with a one year continuation
Sample size:3341
INCLUSION:Patients 18 years of age or older with a diagnosis of depression for which antidepressant treatment was appropriate
EXCLUSION:Previous use of bupropion; pregnant; lactating: anorexic or bulimic; known predisposition to seizures; received an MAO inhibitor within 14 days of the study or an investigational drug within 30 days of the study
OTHER MEDICATIONS/INTERVENTIONS:Other antidepressant medications, neuroleptic drugs, or amphetamine-type drugs were not allowed
POPULATION CHARACTERISTICS:Groups similar at baseline: N/A
Mean age: 43.5
Gender (% female): 59.4
Ethnicity: 96% white; 3% black; 1% other
Other population characteristics:
Psychiatric diagnosis:
Major depression: 73%
Dysthymic disorder: 10%
Bipolar depression: 8%
Atypical depression: 6%
Atypical bipolar: 2%
Other: 1%
OUTCOME ASSESSMENT:Primary Outcome Measures: Number of seizures

Secondary Outcome Measures: N/A

Timing of assessments: Biweekly
RESULTS:
  • Eight seizures were reported in the 3277 patients analyzed during the treatment phase. This is a seizure rate of 0.24%. A survival analysis showed a cumulative seizure rate of 0.36% during the 8 week trial.
ANALYSIS:ITT: No
Post randomization exclusions: N/A
ATTRITION:Overall
Loss to follow-up:NR
Withdrawals due to adverse events:613 (19%)
Withdrawals due to lack of efficacy:NR
Loss to follow-up differential high:N/A
ADVERSE EVENTS:
  • 82 (2.5%) patients experienced major adverse events (life threatening or requiring hospitalization)
  • Most common adverse events were nausea (3.6%), agitation (2.4%), anxiety (1.7%), headache (1.5%), insomnia (1.3%), and rash (1.3%)
QUALITY RATING:N/A
STUDY:Authors: Kasper et al.209
Year: 2009
Country: Multinational
FUNDING:H Lundbeck A/S
DESIGN:Study design: Pooled analysis
Number of patients: 777
AIMS OF REVIEW:To analyze pooled data from two previous studies comparing escitalopram to paroxetine for the long-term treatment of MDD.
STUDIES INCLUDED IN REVIEWTwo double-blinded RCTs comparing escitalopram with paroxetine
TIME PERIOD COVERED:Post-hoc pooled analysis of data from two 6-month RCTs in patients with MDD
CHARACTERISTICS OF INCLUDED STUDIES:RCTs -24-week and 27-week trials -Compared escitalopram to paroxetine
CHARACTERISTICS OF INCLUDED POPULATIONS:Treatment groups had a mean age of 44.6 + or − 13.2 yrs -Baseline MADRS total score of 32.8 + or − 4.7 -Women comprised approx 70% of each group -No significant or clinically relevant differences at baseline between patients treated with escitalopram or paroxetine
CHARACTERISTICS OF INTERVENTIONS:Escitalopram 10–20 mg/d Paroxetine 20–30 mg/d
MAIN RESULTS:See AE
ADVERSE EVENTS:No differences in weight gain between treatment groups -There were no statistically significant differences between treatment groups -Headache and nausea were the most frequent AEs (~20%) -The most common AEs (>10 patients in total) reported during the taper period were: -dizziness (escitalopram 12, paroxetine 15) -headache (escitalopram6, paroxetine 11) -nausea (escitalopram 4, paroxetine 7) -depression (escitalopram 7, paroxetine 4)
COMPREHENSIVE LITERATURE SEARCH STRATEGY:No
STANDARD METHOD OF APPRAISAL OF STUDIES:N/A
QUALITY RATING:N/A
STUDY:Authors: Kennedy SH et al.210
Year: 2006
Country: Canada
FUNDING:Boehringer Ingelheim
DESIGN:Study design: RCT
Setting: Multicenter
Sample size: 141 (131 ITT)
INTERVENTION:
Drug:BupropionParoxetine
Dose:150–300 mg20–40 mg
Duration:8 weeks8 weeks
Sample size:6962
INCLUSION:Outpatients; age 18 – 65 years; DSM-IV criteria for MDD—current MDE of at ≥ 4 weeks. HAM-D ≥ 18; to be in good physical health, sexual interest and activity within the past month; free of any antidepressant use for 2 weeks (4 weeks for fluoxetine)
EXCLUSION:Serious suicide risk; more than 2 failed trials of antidepressant medications at adequate dose and duration during the current episode, drug abuse or dependence within the past 12 months, and a history of bipolar disorder, psychotic disorder, or organic disorder
OTHER MEDICATIONS/INTERVENTIONS:Hypnotic zopiclone (up to 7.5 mg at night) during the first 2 weeks.
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 37.8
Gender (female %): 48
Ethnicity: NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Sexual function Sex FX, IRSD-F
Secondary Outcome Measures: HAM-D
Timing of assessments: Baseline, 2,4,6,8
RESULTS:
  • HAMD Bupropion SR (mean 21.8, SD 2.9) vs. paroxetine (mean 22.2, SD 3.6)
  • HAM-D - men (mean 22.1, SD 3.1) responders 62.9% vs. women (mean 21.9, SD 3.5) responders 53.2%
  • Overall more sexual adverse events with paroxetine than with bupropion
  • No difference between drugs for sexual dysfunction in women
ANALYSIS:ITT: Yes
Post randomization exclusions: 10
ATTRITION:Loss to follow-up: 16% (21) Bupropion 11.6% (8) paroxetine 21% (13)
Withdrawals due to adverse events: NR
Withdrawals due to lack of efficacy: NR
Loss to follow-up differential high: No
ADVERSE EVENTS:
  • None reported
QUALITY RATING:Fair
STUDY:Authors: Khan, et al.211
Year: 2003
Country: US
FUNDING:Not reported
DESIGN:Study design: Meta-analysis
Number of patients: 48,277
AIMS OF REVIEW:Compare suicide rates among depressed patients
STUDIES INCLUDED IN META- ANALYSISPooled analysis of FDA clinical trial data from 1985–2000 for 9 SSRIs
2000 publication reports on 1987 to 1997 (same data)
TIME PERIOD COVERED:1985–2000
CHARACTERISTICS OF INCLUDED STUDIES:FDA clinical trial data
CHARACTERISTICS OF INCLUDED POPULATIONS:Major depression according to DSM-II-R criteria; minimum score of 18 or 20 on HAM-D-17 or HAM-D-21
CHARACTERISTICS OF INCLUDED INTERVENTIONS:Fluoxetine, sertaline, paroxetine, citalopram, fluvoxamine, nefazodone, mirtazapine, buproprion, venlafaxine, imipramine, amitrptyline, maprotiline, trazadone, mianserin, dothiepin
MAIN RESULTS:
  • Absolute Suicide Rate
    • SSRI: 0.15% (0.10–0.20% 95% CI)
    • “Other”: 0.20% (0.09–0.27% 95% CI)
    • Placebo: 0.10% (0.01–0.19% 95% CI)
    • p > 0.05 for difference
  • Suicide Rate by Patient Exposure Years (PEY)
    • SSRI: 0.59%/PEY (0.31–0.87 95% CI)
    • “Other”: 0.76%/PEY (0.49–1.03 95% CI)
    • Placebo: 0.45%/PEY (0.01–0.89 95% CI)
    • p > 0.05 for difference
  • 2000 study: looked at suicide attempts and completion and found no difference
ADVERSE EVENTS:N/A
COMPREHENSIVE LITERATURE SEARCH STRATEGY:No
STANDARD METHOD OF APPRAISAL OF STUDIES:Not reported
QUALITY RATING:Fair
STUDY:Authors: Kharofa J et al212
Year: 2007
Country: USA
FUNDING:None
DESIGN:Study design: Case-control study
Setting: Emergency rooms and hospitals
Sample size: 916
Cases: patients with intracerebral (ICH) and subarachnoid hemorrhage (SAH) on citalopram, escitalopram, fluoxetine, paroxetine, and sertraline.Controls: matched patients on citalopram, escitalopram, fluoxetine, paroxetine, and sertraline
Sample size:9161776
INCLUSION:Cases of intracerebral (ICH) and subarachnoid hemorrhage (SAH) were identified in the Greater Cincinnati region
EXCLUSION:NR
OTHER MEDICATIONS/INTERVENTIONS:Warfarin Cases 77 (8.4%) Controls 43 (2.4%)
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 57.3
Gender (female %): NR
Ethnicity: NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Hemorrhagic stroke
Timing of assessments: May 1997 to August 2001 and from July 2002 to October 2005
RESULTS:Of the 916 hemorrhagic stroke patients, 71 (7.8%) were on an SSRI at the time of stroke, and of 1776 demographically matched controls, 158 (8.9%) were on an SSRI. After controlling for multiple risk factors, SSRI use was not independently associated with increased risk for hemorrhagic stroke (OR = 0.8, 95% CI: 0.5 to 1.2; P = 0.25).
ANALYSIS:ITT: NA
Post randomization exclusions: NA
Loss to follow-up: NA
ATTRITION:NA
Withdrawals due to adverse events:
Withdrawals due to lack of efficacy:
Loss to follow-up differential high:
ADVERSE EVENTS:
  • See results
QUALITY RATING:Fair
STUDY:Authors: Kiev, et al.55
Year: 1997
Country: US
FUNDING:Solvay Pharma, Upjohn
DESIGN:Study design: RCT
Setting: Single center
Sample size: 60
INTERVENTION:
Drug:FluvoxamineParoxetine
Dose:50–150 mg/d20–50 mg/d
Duration:7 weeks7 weeks
INCLUSION:Age 18–65; meet DMS-III-R criteria for single or recurrent MDD; ≥ 20 on HAM-D-21 (including minimum score of 2 on depressed mood item)
EXCLUSION:Non-English speakers; history of medication non-compliance; demonstration of placebo response during run-in, history of substance abuse; severe suicide risk or auto-aggressive behavior; used a drug within 30 days with anticipated major organ toxicity; pregnancy, lactation; hypersensitivity to SSRIs; participation in prior drug 1 studies; other significant organic disease; clinically significant lab abnormalities; other primary psychiatric diagnoses; transportation difficulties
OTHER MEDICATIONS/INTERVENTIONS:Antacids, laxatives, acetaminophen, aspirin, ibuprofen, chloral hydrate, other meds only with permission of study physician
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: Fluvoxamine: 42.7, paroxetine: 39
Gender (female%): Fluvoxamine: 53%, paroxetine: 53%
Ethnicity: White: fluvoxamine: 87%, paroxetine: 93%
Other population characteristics: Not reported
OUTCOME ASSESSMENT:Measures: HAM-D-21, HAM-A, SCL-56, CGI
Timing of assessments: Baseline, weeks 1, 2, 3, 5, 7
RESULTS:
  • Mean change in HAM-D score: fluvoxamine: −13.45, paroxetine: −12.86 (p = 0.763)
  • No significant differences between groups on HAM-D-21, CGI, HAM-A, or SCL56
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes
ATTRITION:Loss to follow-up: 31%
Withdrawals due to adverse events: fluvoxamine: 6.8%, paroxetine: 13.8%
Loss to follow-up differential high: No
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Landen M, et al.213
Year: 2005
Country: Sweden and Norway
FUNDING:Bristol-Myers Squibb, Sweden
OBJECTIVE:To determine: 1) concordance of sexual dysfunction adverse event rates between open-ended questioning and directed questioning; 2) the incidence of sexual side effects of citalopram and paroxetine; 3) the correlation between sexual side effects and illness severity, treatment duration and drug/dose combination
DESIGN:Study design: Non-randomized trial of adverse event elicitation methods embedded in a RCT (Landen et al 1998 – patients who had not responded to CP or PX were randomized to receive buspirone or placebo)
Setting: Multi-center (13 centers)
Sample size: 119
INTERVENTION:
Drug:CitalopramParoxetine
Dose:at least 40 mg/dat least 30 mg/d
Duration:4 weeks4 weeks
Sample size:7742
INCLUSION:Patients 18 years or older; met criteria for a major depressive episode according to DSM-IV criteria; has not responded to CP or PX for a minimum of 4 weeks prior to start of study
EXCLUSION:Pregnancy; epilepsy; severe somatic disease; mental disorder due to a general medical condition; substance abuse; highly suicidal status
OTHER MEDICATIONS/INTERVENTIONS:Patients received either buspirone or placebo for 4 week study duration
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 46
Gender (% female): 69%
Ethnicity: NR
Other population characteristics: NR
OUTCOME ASSESSMENT:Primary Outcome Measures: Sexual dysfunction score (0–6); Percent patients reporting any sexual side effect based on open and direct questioning

Secondary Outcome Measures: N/A

Timing of assessments: Before and after the 4 week trial
RESULTS:By objective
  1. Side effect elicitation method
    • Significantly more patients (49 versus 6) reported sexual side effects in response to direct questioning than open questioning (p < 0.001).
  2. Incidence of side effects by drug
  3. Correlations with illness severity and treatment parameters
    • Only weak correlation with duration of current depression episode (p = 0.043)
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Withdrawals due to lack of efficacy: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:
  • Decreased desire reported by 43% of men and 32% of women
  • Orgasmic dysfunction reported by 23% women and 32% men
QUALITY RATING:Good
STUDY:Authors: Lopez-Ibor JJ214
Year: 1993
Country: Spain
FUNDING:NR
DESIGN:Study design: Retrospective database analysis
Setting: Not reported
Sample size: 4,668
INTERVENTION:
Drug:ParoxetinePlaceboActive control
Dose:Not reportedN/AN/A
Duration:Up to 6 weeksUp to 6 weeksUp to 6 weeks
INCLUSION:Depressed patients enrolled in a clinical trial
EXCLUSION:Not reported
OTHER MEDICATIONS/ INTERVENTIONS:Not reported
POPULATION CHARACTERISTICS:Groups similar at baseline: Not reported
Mean age: Not reported
Gender: Not reported
Ethnicity: Not reported
Other population characteristics: Not reported
OUTCOME ASSESSMENT:Measures: Suicide item of HAM-D, emergence of suicidal ideation, assessed by the development of HAM-D suicide item score
Timing of assessments: N/A
RESULTS:Paroxetine and active control were significantly better than placebo in reducing suicidal thoughts and behavior from week 1 onwards
ANALYSIS:ITT: N/A
Post randomization exclusions: Not reported
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:
  • There were no differences among the groups with regards to suicidality as an adverse event.
  • 0.4% of each group reported suicidality.
  • There were 10 suicides overall and 58 attempts overall.
QUALITY RATING:N/A
STUDY:Authors: Mackay, et al.215, 216
Year: 1997
Country: UK
FUNDING:Drug Safety Research Unit, UK, various unnamed pharmaceutical companies
DESIGN:Study design: Cohort study (prescription event monitoring)
Setting: General practice in the UK
Sample size: Number identified as getting a first prescription” fluvoxamine: 20,504, fluoxetine: 24,738, sertraline: 24,632, paroxetine: 26,194
INTERVENTION:
Drugs:Drugs compared: fluvoxamine, fluoxetine, sertraline, paroxetine
Dose:N/A
Duration:Outcomes assessed after approximately 6 months for all but fluovoxamine (which was 12 months)
INCLUSION:Patients who received a first prescription from their GP during the following time periods: fluvoxamine: Feb 1987 – Feb 1988; fluoxetine: Mar 1989 – Mar 1990; sertraline: Jan 1991 – Sep 1992; paroxetine: Mar 1991 – Mar 1992
EXCLUSION:Not reported
OTHER MEDICATIONS/ INTERVENTIONS:Not reported
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes; some differences existed between groups as far as indication for prescription
Mean age: 50
Gender (% female): 70%
Ethnicity: Not reported
Other population characteristics: Not reported
OUTCOME ASSESSMENT:Measures: GP completion of a simple questionnaire (green form), questions asked: perceived efficacy, reason for stopping, indication for prescribing, duration of therapy, and events during and after treatment. (Event = new diagnosis, reason for referral to a consultant or admission to hospital, unexpected deterioration (or improvement) in a concurrent illness, suspected drug reaction or any complaint which was considered of sufficient importance to enter in patient notes.
Timing of assessments: Mailed 6–12 months after initial prescription written
RESULTS:
  • Reasons for discontinuation in 1st month of treatment due to adverse events:
Incidence Densities (Events/1000 patient-months)
FluvoxamineFluoxetineSertralineParoxetine
Nausea/vomiting127.226.334.652.9
Malaise/lassitude41.516.312.017.8
Drowsiness/sedation*22.68.27.320.5
Dizziness25.56.78.711.5
Headache/migraine25.113.513.113.1
Tremor*13.25.76.212.4
* (p < 0.001 for fluoxetine and sertraline vs. fluvoxamine and paroxetine)
  • Adverse Effects Reported:
Incidence Densities (Events/1000 patient-months)s
FluvoxamineFluoxetineSertralineParoxetine
Nausea/vomiting42.89.08.613.0
Malaise/lassitude15.25.53.75.2
Dizziness9.62.72.84.0
Headache/migraine10.15.75.44.8
Mean17.67.06.24.8
  • No statistical differences in onset of mania or hypomania with any of the SSRIs
  • No serious cardiac events with any of the SSRIs
  • No deaths attributed to SSRIs. No difference in the number of suicides with each of the four SSRIs (approx 0.2–0.3% in each arm)
SSRIs and nefazodone:
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Completion rates of surveys: 60%
Withdrawals due to adverse events: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:N/A
QUALITY RATING:Fair
STUDY:Authors: Maina G, et al.217
Year: 2004
Country: Italy
FUNDING:None
DESIGN:Study design: Non-randomized, open-label trial
Setting: Single center (Department of Neuroscience, University of Turin)
Sample size: 149 started trial
INTERVENTION:
Drug:ClomipramineCitalopramFluoxetineParoxetineFluvoxamineSertraline
Dose:150–250 mg/d40–80 mg/d40–80 mg/d40–80 mg/d200–300 mg/d150–200 mg/d
Duration:2.5 years2.5 years2.5 years2.5 years2.5 years2.5 years
Sample size:232123212822
INCLUSION:Patients 18 years of age or older; Met DSM-IV criteria for OCD based on the Structured Clinical Interview; YBOCS score greater than or equal to 16; completed 6 month acute treatment phase of trial; gave informed consent
EXCLUSION:Pregnant; lactating; current or past diagnosis of eating disorder, schizophrenia, or other psychotic disorders; organic mental disorder; medical illness; met diagnostic criteria for a major depressive episode; had a HAM-D17 score greater than or equal to 15
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 34.9 years
Gender: 51% female
Ethnicity: NR
Other population characteristics:
  • Mean duration of illness: 12.1 years
OUTCOME ASSESSMENT:Primary Outcome Measures: Percentage weight gain

Secondary Outcome Measures: Number of patients with extreme weight gain

Timing of assessments: Weight recorded at the beginning of treatment and at six months intervals thereafter.
RESULTS:
ANALYSIS:ITT: No
Post randomization exclusions: N/A: above results are reported only for patients who completed the 2 year extension phase of the trial
ATTRITION:Loss to follow-up: 7%
Withdrawals due to adverse events: NR
Loss to follow-up differential high: NR
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: March JS116–120, 218
Year: 2004, 2006, 2009
Country: US
Trial name: TADS
FUNDING:NIMH
DESIGN:Study design: RCT
Setting: Multi-center (13 sites-academic and community clinics)
Sample size: 439
INTERVENTION:[blinded][blinded][unblinded][unblinded]
Drug:PlaceboFluoxetineFluoxetine and CBTCBT alone
Dose:N/A10–40 mg/d10–40 mg/dN/A
Duration:12 weeks12 weeks12 weeks12 weeks
Sample Size:112109107111
INCLUSION:Ages 12–17; ability to receive care as an outpatient; a DSM-IV diagnosis of MDD at consent and again at baseline; a CDRS-R total score of 45 or higher at baseline; a full scale IQ of 80 or higher; not taking antidepressants prior to consent; depressive mood present in at least 2 or 3 contexts (home, school, among peers) for a least 6 wks prior to consent
EXCLUSION:Current or past diagnosis of bipolar disorder, severe conduct disorder, current substance abuse or dependence; pervasive developmental disorders, thought disorder; concurrent treatment with psychotropic medication or psychotherapy outside the study; 2 failed SSRI trials; a poor response to clinical treatment containing CBT for depression; intolerance to fluoxetine; confounding medical condition, non-English speaking patient or parent; pregnancy or refusal to use birth control; suicidal in the past 6 months; patients considered to be a danger to themselves or others
OTHER MEDICATIONS/ INTERVENTIONS:Concurrent stable psychostimulant treatment (methylphenidate or mixed amphetamine salts) for attention deficit hyperactivity disorder permitted
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 14.6 (treatment-specific numbers not reported)
Gender (% female): 54.4% (treatment-specific numbers not reported)
Ethnicity: White: 73.8%; black: 12.5%; Hispanic: 8.9% (treatment-specific numbers not reported)
Other population characteristics: None significant
OUTCOME ASSESSMENT:Measures: CDRS-R total score; CGI-I; RADS; SIQ-Jr, Functioning: Children’s Global Assessment Scale (CGAS), global health with the Health of the Nation Outcome Scales for Children and Adolescents (HoNOSCA), and quality of life with the Pediatric Quality of Life Enjoyment and Satisfaction Questionnaire (PQ-LES-Q)
Timing of assessments: Baseline and weeks 6 and 12
RESULTS:
  • Fluoxetine with CBT was statistically significantly better than placebo (p = 0.001) on the CDRS-R
  • Compared to fluoxetine alone (p = 0.02) and CBT alone (p = 0.01), treatment with fluoxetine and CBT was statistically significantly superior on the CDRS-R
  • Fluoxetine alone was superior to CBT alone (p = 0.01) on the CDRS-R
  • Fluoxetine with CBT (p < 0.001) and fluoxetine alone (p < 0.001) demonstrated significant improvement on the CGI-I compared to placebo; CBT alone was not significantly better than placebo (p = 0.20)
  • Fluoxetine plus CBT were significantly better than placebo, fluoxetine alone, or CBT alone (p < 0.01) on the RADS
  • Clinically significant suicidal thinking improved significantly in all four treatment groups (SIQ-Jr), with fluoxetine plus CBT showing the greatest reduction (p = 0.02)
  • Loss of MDD diagnosis (using DSM-IV, K-SADS-P/L) at week 12: Both fluoxetine (78.6%) and fluoxetine+CBT(COMB) (85.3%) were superior to CBT alone (61.1%) and placebo (60.4%).
  • Remission rate (CDRS-R≤28): COMB was superior to all other groups (COMB 37% vs. FLX 23% vs. CBT 16% vs. PBO 17%)
  • Response rate (CGI-I≤2): COMB 71.0% vs. FLX 43.2% vs. CBT 43.2% vs. PBO 34.8%
  • Functioning and QOL: COMB was better than placebo on all measures, and better then FLX on CGAS and PQ-LES-Q. Fluoxetine was superior to both placebo and CBT on the CGAS only. CBT monotherapy was not statistically different from the placebo group on any of the measures assessed. The combination of fluoxetine and CBT was effective in improving functioning, global health, and quality of life in depressed adolescents. Fluoxetine monotherapy improved functioning.
  • LONG-TERM: 327 patients completed 36 weeks (after 12 weeks an open trial, no placebo). By week 24 all treatments converged, and remained so to 36 weeks (response rates COMB 86% vs. FLX 81% vs. CBT 81%).
  • Risk of suicidality does not increase over time
  • No difference in event timing (suicidal event) for patients receiving medication versus those not on medication. (events occurred 0,4–31,1 weeks [mean 11,9 +/−8,2] after starting TADS treatment
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes
ATTRITION:Loss to follow-up: 18.2%; fluoxetine+CBT: 14%; fluoxetine: 17%; CBT: 22%; placebo: 21%
Withdrawals due to adverse events: Not reported
Loss to follow-up differential high: No
ADVERSE EVENTS:Adverse events reported as harm-related, psychiatric, or other
QUALITY RATING:Good
STUDY:Authors: Martinez C, et al.219
Year: 2005
Country: UK
FUNDING:Medicines and Healthcare products Regulatory Agency
DESIGN:Study design: Case control study
Setting: General Practice Research Database (clinical primary care records in the UK )
Sample size: 146,095
INTERVENTION:Cases (suicide and non-fatal self-harm)Controls
Drug:SSRIs/TCAsSSRIs/TCAs
Dose:NRNR
Duration:1995–20011995–2001
Sample size (suicides/self-harm):2037 (69/1968)35,615
INCLUSION:Individuals 90 years or younger with a first prescription for antidepressants between January 1, 1995 and December 31, 2001 entered in the General Practice Research Database; diagnosed with depression
EXCLUSION:None
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 31% of patients were in the age cohort 31–45 years old
Gender: 65% female
Ethnicity: NR
Other population characteristics:
  • History of self harm: <1 % patients
OUTCOME ASSESSMENT:Primary Outcome Measures: Risk of non-fatal self harm and completed suicide

Secondary Outcome Measures: none

Timing of assessments: N/A
RESULTS:
  • No difference in risk of non-fatal self harm among the different SSRIs (p =0.35). The greatest risk of self harm was found in patients taking paroxetine.
  • No difference in the risk of self-harm between SSRIs and TCAs (OR: 0.99 CI: 0.86 to 1.14).
  • Significantly higher risk of self-harm among SSRI patients younger than 18 years compared to those on TCAs (OR 1.59; 95% CI 1.01–2.50). Among SSRIs, the greatest risk of self harm was found in patients taking paroxetine.
  • No difference in the risk of suicide between SSRIs and TCAs (OR: 0.57 CI: 0.26 to 1.25).
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:N/A
QUALITY RATING:Good
STUDY:Authors: Martinez et al.220
Year: 2010
Country: UK
FUNDING:Wyeth Inc
DESIGN:Study design: Nested case-control analysis
Setting: United Kingdom General Practice Research Database (GPRD).
Sample size: 207384
INTERVENTION:Cases (sudden cardiac death or near death)Controls
Drug:Venlafaxine, fluoxetine, citalopram, dosulepinVenlafaxine, fluoxetine, citalopram, dosulepine
Dose:VariousVarious
Duration:Mean 3.3 yearsMean 3.3 years
Sample size:56814812
INCLUSION:New users of venlafaxine, fluoxetine, citalopram, or dosulepin on or after 1 January 1995, aged 18 to 89 years with a diagnosis of depression or anxiety.
Patients were included if they had a permanent registration status with a participating general practice, had at least one year longitudinal record before the incident prescription, had an acceptable patient status for data quality, and originated from a general practice which was up to standard for at least a year before the incident prescription.
EXCLUSION:Patients with a history of life threatening ventricular tachyarrhythmia, cardioversion, aborted cardiac arrest, or implantation of a cardiac defibrillator before cohort entry were excluded. Patients with a congenital conduction disorder or advanced cardiomyopathy (hypertrophic or dilated) before cohort entry or at any time during follow-up were also excluded
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes – but more alcohol abuse (2.8 vs. 1) and smokers in cases (43.8 vs. 37.3) P = NR
Mean age: 72.9 years
Gender (female %): 54.6
Ethnicity (Caucasian %): NR
Other population characteristics: more alcohol abuse (2.8 vs. 1) and smokers in cases (43.8 vs. 37.3) P = NR
OUTCOME ASSESSMENT:Primary Outcome Measures: risk of sudden cardiac death or near death (identified from medical records indicating non-fatal acute ventricular tachyarrythmia, sudden death due to cardiac causes, or out of hospital deaths from acute ischaemic events)
Secondary Outcome Measures:
Timing of assessments: NA
RESULTS:Risk of sudden cardiac death or near death associated with venlafaxine use was 0.66 (95% CI 0.38 to 1.14) relative to fluoxetine use, whereas compared with citalopram it was 0.89 (0.50 to 1.60)
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS: see results
QUALITY RATING:Fair
STUDY:Authors: Meijer WE, et. al.221
Year: 2002
Country: The Netherlands
FUNDING:Pfizer
DESIGN:Study design: Observational study of adverse effects
Setting: Multi-center (109 psychiatrists)
Sample size: 1,251
INTERVENTION:
Drug:Observed: Sertraline or fluoxetine, fluvoxamine, or paroxetine
Dose:Any administered dose
Duration:12 month observation period
INCLUSION:All patients with a new sertraline prescription; patients taking fluoxetine, fluvoxamine, or paroxetine were used as controls
EXCLUSION:None reported
ALLOWED OTHER MEDICATIONS/ INTERVENTIONS:None reported
POPULATION CHARACTERISTICS:Groups similar at baseline: N/A
Mean age: 41
Gender (% female): 64.1%
Ethnicity: Not reported
Other population characteristics: Significantly more sertraline patients had a diagnosis of depressive disorder than patients on other SSRIs (p < 0.001); anxiety disorder was significantly less in sertraline patients than patients with other SSRIs (p < 0.001); MDD: 77.9%, anxiety: 15.5%, multiple diagnoses: 37.8%.
OUTCOME ASSESSMENT:Measures: Physicians recorded adverse events at each patient visit, used WHO coding; serious adverse events (SAEs) recorded according to the International Conference on Harmonization of Good Clinical Practice (ICH-CGP)
Timing of assessments: Not reported
RESULTS:
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:N/A
QUALITY RATING:Fair
STUDY:Authors: Montejo et al.222
Year: 2001
Country: Spain
FUNDING:Bristol-Myers Squibb
DESIGN:Study design: Observational
Setting: Multi-center
Sample size: 1022
INTERVENTION:
Drug:fluoxetineparoxetinefluvoxaminesertralinecitalopramvenlafaxinemirtazapinenefazodone
Dose (mean):24.5 mg23.4 mg115.7 mg90.4 mg28.7 mg159.5 mg37.7 mg324.6 mg
Duration:NRNRNRNRNRNRNRNR
Sample size:2792087715966554950
INCLUSION:Normal sexual functioning prior to taking antidepressants; treatment with an antidepressant alone or in combination with a benzodiazepine; previous regular and satisfactory sexual practices; occurrence of sexual dysfunction within the two months after introduction of an antidepressant
EXCLUSION:Prior sexual dysfunction; combination of antidepressant and neuroleptic treatment; treatment with hormones or any other drug capable of interfering with sexual intercourse; significant intercurrent diseases affecting sexual function; substance abuse
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: NR
Mean age: Overall: 39.8
Gender (% female): Overall: 60%
Ethnicity: NR
Other population characteristics: MDD: 60.1%; dysthymic disorder: 17.3%; panic disorder: 12.1%; OCD: 5.9%; other disorders: 3.7%
OUTCOME ASSESSMENT:Primary Outcome Measures: PRSexDQ (Pscychotropic-Related Sexual Dysfunction Questionnaire)

Secondary Outcome Measures: None

Timing of assessments: Each clinic visit
RESULTS:
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Loss to follow-up: N/A
Withdrawals due to adverse events: N/A
Withdrawals due to lack of efficacy: N/A
Loss to follow-up differential high: N/A
ADVERSE EVENTS:N/A
QUALITY RATING:Fair
STUDY:Authors: Nierenberg A, et al. 69 Pigott T, et al.70 and Clayton A, et al.71
Year: 2007
Country: USA
FUNDING:Eli Lilly Inc
DESIGN:Study design: RCT
Setting: Multicenter
Sample size: 684 (114 for Clayton subanalysis of CSFQ)
INTERVENTION:
Drug:DuloxetineEscitalopramPlacebo
Dose:60 mg10 mgNA
Duration:8 weeks and 8 months8 weeks and 8 months8 weeks and 8 months
Sample size:273274137
INCLUSION:18 years old; diagnosed with MDD; MADRS > 22 and CGI-S > 4; normal or clinically unremarkable exam, lab and ECG
EXCLUSION:Pregnant, lactation; primary Axis 1 disorder other than MDD;; previous diagnosis bipolar, schizophrenia or other psychotic disorders or Axis 2 disorder that might interfere; significant risk of suicide; substance dependence; treatment resistant; ECT.
OTHER MEDICATIONS/ INTERVENTIONS:Chronic use of certain prescriptions such as ACE inhibitors, alpha and beta blockers, anti-arrhythmics, and calcium channel blockers if on stable dose for at least 3 months
POPULATION CHARACTERISTICS:Groups similar at baseline: No
Mean age: Duloxetine 41.1 escitalopram 43.3 placebo 42.5
Gender (female %): overall 65.2% duloxetine 63.4% escitalopram 67.9% placebo 63.5%
Ethnicity: Overall 77.6% Caucasian Duloxetine 75.5% escitalopram 77.4% placebo 82.5%
Other population characteristics: Mean HAM-D Duloxetine 17.6 escitalopram 17.8 placebo 17.7
OUTCOME ASSESSMENT:Primary Outcome Measures: Onset of efficacy HAM-D at 8 months and CSFQ
Secondary Outcome Measures: HAM-D, HAM-A, CGI-S
Timing of assessments: Baseline, weeks 1,2,3,4,6,8
RESULTS:
  • Mean change Duloxetine vs. escitalopram v. placebo 8 weeks and 8 months
  • HAM-D −7.61 (0.42) vs. −7.22 (0.40) vs. −5.97 (0.58) P < 0.05 Duloxetine vs. placebo and −10.55 (0.48) vs. −10.91 (0.45) vs −8.06 (1.13)
  • CGI-S −1.44 (0.08) vs. 1.36(0.07) vs. −1.08 (0.11) P < 0.01 Duloxetine vs. placebo and P < 0.05 Escitalopram vs. placebo and −2.17 ((0.09) vs. −2.20 (0.09) vs. −2.11 (0.22)
  • HAM-A −5.49 (0.36)) vs. −5.16 (0.34) vs. −4.32 (0.50) and −7.30 (0.44) vs. −7.92 (0.41) vs. −5.73 (1.03)
  • Response HAM-D 48.7% vs. 45.3% vs. 36.9%
  • Remission HAM-D 37% vs. 32% vs. 27% and 70% vs. 75% vs. NR
  • 8 week incidence of treatment-emergent sexual dysfunction duloxetine 17/51 (33.3%) escitalopram; 19/39 (48.7%) placebo 4/24 (16.7%) (P = 0.01 escitalopram vs. placebo; P = 0.13 duloxetine vs. placebo) and at 8 months duloxetine 33.3% escitalopram 43.6% placebo 25%
ANALYSIS:ITT: Yes
Post randomization exclusions:
ATTRITION:Loss to follow-up: Duloxetine 85, escitalopram 66, placebo 40
Withdrawals due to adverse events: Duloxetine 20, escitalopram 14, placebo 8
Withdrawals due to lack of efficacy: Duloxetine 9, escitalopram 4, placebo 7
Loss to follow-up differential high: No
ADVERSE EVENTS:
  • Duloxetine vs. escitalopram v. placebo (%) 8 weeks and 8 months
  • Nausea 23.8* ** vs. 12.0 vs. 8.8 and 29.3* vs. 14.2 vs. 10.2
  • Dry mouth 21.6* ** vs. 10.9 vs. 10.9 and 24.2* ** vs. 11.7 vs. 11.7
  • Headache 19.4 vs. 20.1 vs. 14.6 and 25.6* vs. 23.7 vs. 16.1
  • Diarrhea 11.7 vs. 12.0 vs. 8.0 and 13.2 vs. 17.5* vs.9.5
  • Dizziness 9.5 vs. 7.3 vs. 5.1 and 12.5 vs. 11.7 vs. 7.3
  • Constipation 8.4 vs. 5.8 vs. 5.8 and 11.0 vs. 8.4 vs. 6.6
  • Decreased appetite 8.1* vs. 4.7 vs. 2.2 and 8.1* vs. 5.1 vs. 2.2
  • Insomnia 8.1 vs. 7.7 vs. 6.6
  • Hyperhidrosis* 7.7 vs. 4.0 vs. 0.7 and 9.9* vs. 5.5 vs. 1.5
  • Vomiting 7.3* ** vs. 2.2 vs. 0.7 and 9.2* ** vs. 3.6 vs. 1.5
  • Somnolence 5.9 vs. 6.6 vs. 3.6 and 7.3 vs. 7.3 vs. 4.4
  • Nasopharyngitis 5.5 vs. 6.6 vs. 6.6 and 8.4 vs. 10.9 vs. 8.0
  • Yawning 5.5* ** vs. 2.2 vs. 0 and 5.9* ** vs. 2.2 vs. 0
  • Decreased libido 5.1 vs. 4.0 vs. 2.2 and 6.6 vs. 6.6 vs. 2.9
  • Fatigue 5.1 vs. 6.2 vs. 8.0 and 8.1 vs. 9.9 vs. 8.8
  • Anxiety 4.4 vs. 2.9 vs. 5.8 and 5.5 vs. 3.6 vs. 5.8
  • Back pain NR and 5.5 vs. 5.5 vs. 3.6
  • Dyspepsia NR and 5.9 vs. 4.7 vs. 4.4
  • Anthralgia NR and 4.0 vs. 5.1 vs. 3.6
  • Blurred vision NR and 5.9 vs. 3.3 vs. 2.2
  • Anorgasmia NR and 4.8* vs. 4.0 vs. 0
  • Pain in extremity NR and 3.7 vs. 4.7* vs. 0.7
  • Increased weight NR and 2.6 vs. 5.5* vs. 0
  • Abnormal dreams NR and 4.8* vs. 1.8 vs. 0.7
  • Sedation NR and 4.0* vs. 1.8 vs. 0
  • Night sweats NR and 3.7** vs. 0 vs. 0.7
  • Migraine NR and 0.4 vs. 2.9** vs. 0.7
  • * P < 0.05 vs. placebo and ** P < 0.05 duloxetine vs. escitalopram
QUALITY RATING:Fair
STUDY:Authors: Nieuwstraten C, et al.73
Year: 2001
Country: Canada
FUNDING:Not reported
DESIGN:Study design: Meta-analysis
Number of patients: 1332
AIMS OF REVIEW:To assess the benefits and risks of bupropion vs. SSRIs in major depression
STUDIES INCLUDED IN META- ANALYSISKavoussi RJ et al. 1997, Segraves RT, et al. 2000, Weihs KL, et al. 2000, Croft H, et al. 1999, ColemanCC, et al. 1999, Feighner JP, et al. 1991
TIME PERIOD COVERED:1966–1999
CHARACTERISTICS OF INCLUDED STUDIES:RCTs, study durations: 6–16 weeks, median 7 weeks
CHARACTERISTICS OF INCLUDED POPULATIONS:Age: 36 to 70 yrs; proportion of females: 48.0% to 61.8%
CHARACTERISTICS OF INCLUDED INTERVENTIONS:Bupropion vs. sertraline (3 trials), bupropion vs. paroxetine (1 trial), bupropion vs. fluoxetine (1 trial)
MAIN RESULTS:Results of HAM-D scores and CGI-I scores could not be pooled due to the unavailability of data; the weighted mean differences of CGI-S and HAM-A scores were not significantly different between bupropion and SSRIs
ADVERSE EVENTS:Nausea, diarrhea, and somnolence occurred significantly less frequently in the bupropion group compared to the SSRI group RR: nausea: 0.6 (95%CI: 0.41–0.89), diarrhea: 0.31 (95%CI: 0.16–0.57), somnolence: 0.27 (95%CI: 0.15–0.48). Satisfaction with sexual function was significantly less in the SSRI group RR: 1.28 (95%CI: 1.16–1.41)
COMPREHENSIVE LITERATURE SEARCH STRATEGY:Yes
STANDARD METHOD OF APPRAISAL OF STUDIES:Yes
QUALITY RATING:Good
STUDY:Authors: Olfson et al. 223
Year: 2008
Country: US
FUNDING:NARSAD, AHRQ, American Foundation for Suicide Prevention
DESIGN:Study design: case-control study
Setting: outpatient/ community
Sample size: 1368
INTERVENTION: = exposureCases (suicide attempt)Controls (no suicide attempt)
Drug:Exposure to any antidepressant (subclassified in SSRIs or any other antidepressant)Exposure to any antidepressant (subclassified in SSRIs or any other antidepressant)
Dose:No data shown (dosage rating based on 4-point scale, from 1= low, to 4= high)No data shown (dosage rating based on 4-point scale, from 1= low, to 4= high)
Duration:No data shown (dichotomized as < 30 days or ≥ 30 days before event date)No data shown (dichotomized as < 30 days or ≥ 30 days before event date)
Sample size:2361132
INCLUSION:Outpatients with new episodes of treatment of major depressive episode, ages 6–64 years with at least one claim in the Medicaid administrative database during January 1, 1999 through December 31, 2000 (and Medicaid eligibility 90 before and 120 days after index episode, respectively); treatment vs. no treatment with antidepressants (dichotomized as SSRIs including fluoxetine, fluvoxamine, paroxetine, sertraline, citalopram, escitalopram and other antidepressant)
EXCLUSION:Previous episodes of MDD (evidenced as any diagnostic claim of MDD or claims of electroconvulsive therapy or antidepressant/ antipsychotic medication or treatment with mood stabilizer within 90 days before index episode); previous suicide attempt (90 days before index episode of MDD); pregnancy; mental/ psychiatric psychotic disorders
OTHER MEDICATIONS/ INTERVENTIONS:Psychotherapy
POPULATION CHARACTERISTICS:Groups similar at baseline: matched on age (within +/− 3 years), sex and ethnicity (white non-white); data for comparison of cases and controls not shown
Mean age: children: 15.1 yrs (± 1.4); adults: 31.6 ± (10.1)
Gender (female %): children: 80.4%; adults: 68.3%
Ethnicity (white, non-Hispanic %): children: 76.5%; adults: 78.9%
Cases and controls matched on: depression severity, recent treatment of substance use disorder, other depression-related disorder, major depressive disorder type, recent treatment with psychotherapy (data not shown for cases and controls but reported as matched on all these criteria)
OUTCOME ASSESSMENT:Primary Outcome Measures: ICD-9 (CM 950–959) coded suicide attempts, including all types of intentional self-injury; subclassified by major self-injury category (drug ingestions, cutting, all other types); conditional OR of suicide attempt, separately analyzed for children & adolescents (6–18 years) and adults (19–64 years); within each age group separately for both sex groups, and for all 4 groups of depression type and 3 groups of depression severity; all analyses adjusted for duration and dosage of antidepressant, any psychotherapy
Secondary Outcome Measures: NA
Timing of assessments: NA (secondary analysis; duration dichotomized as < 30 days or ≥ 30 days before event date)
RESULTS:Among adults risk of suicide attempt not significantly associated with antidepressant use [OR= 0.85; 95% CI (0.57 – 1.28) P = 0.44; cases, N= 185; controls, N= 893], but among adult males statistically significant protective effect [OR= 0.32; 95% CI (0.12 – 0.83), P = 0.01; cases, N= 57; controls, N= 268]; among children statistically significant association of antidepressant use and suicide attempts [OR= 2.08; 95% CI 1.06 – 4.10) P = 0.03; cases, N= 51; controls, N= 239]; SSRIs as a class (including fluoxetine, fluvoxamine, paroxetine, sertraline, citalopram, escitalopram) not statistically significant; no statistical significant association of severity of disease, type of depressive disorder or recent psychotherapy visit;
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS:See results
QUALITY RATING:Good
STUDY:Authors: Pedersen AG224
Year: 2005
Country: Multinational
FUNDING:H. Lundbeck A/S
DESIGN:Study design: Retrospective cohort study
Setting: Clinical trials
Sample size: 4,091
INTERVENTION:
Drug:EscitalopramPlacebo
Dose:5–20 mg/dayN/A
Duration:8–24 weeks8–24 weeks
Sample size:26481443
INCLUSION:Adult outpatients with MDD (2277) or anxiety (371)
EXCLUSION:NR
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: NR
Mean age: NR
Gender (% female): NR
Ethnicity: NR
Other population characteristics: NR
OUTCOME ASSESSMENT:Primary Outcome Measures: Rates of suicide and self-harm

Secondary Outcome Measures:

Timing of assessments: N/A
RESULTS:
ANALYSIS:ITT: N/A
Post randomization exclusions: N/A
ATTRITION:Overall
Loss to follow-up: NR
Withdrawals due to adverse events: NR
Withdrawals due to lack of efficacy: NR
Loss to follow-up differential high: Not enough information
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Rahme et al. 225
Year: 2008
Country: Canada
FUNDING:Researchers funded by: Fonds de la Recherche en Sante du Quebec, Canadian Institutes of Health
One researcher is consultant for Pfizer Canada
DESIGN:Study design: retrospective cohort study
Setting: N.A
Sample size: N.A
INTERVENTION:
Drug:SSRIs (citalopram, fluoxetine, fluvoxamine, paroxetine, sertraline)
Dose:N.A
Duration:at least 365 days
Sample size:128,229 patients during periods of use and no use of antidepressants
INCLUSION:Patients 65 years and older who had filled a prescription for an SSRI between January 1998 and December 2004, whose data were available from the Quebec Health Care Fund and Vital Statistics databases.
EXCLUSION:NA
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 75.4 years
Gender (female %): 70%
Ethnicity (Caucasian %): NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures:
Number of suicide deaths (crude rate/100.000 patient-years) during SSRI (citalopram, fluoxetine, fluvoxamine, paroxetine, sertraline) use, other antidepressant use, during the use of both, an SSRI and another antidepressant, during no antidepressant use.
Adjusted risk of suicide death during SSRI use versus nonuse
Adjusted risk of poisoning during SSRI use versus nonuse
Secondary Outcome Measures:
Timing of assessments:
RESULTS:
Numbers of suicide deaths (crude rate/100.000 patient-years):
During SSRI use: 37
During other antidepressant use: 16
During use of both an SSRI and another antidepressant: 5
During no antidepressant use: 29

Adjusted risk of suicide death (Cox regression models with time-dependent exposure):
  • The risk of suicide death during antidepressant treatment overall was not higher than during times without treatment; Hazard ratio (HR): 0.84 (95% CI 0.52–1.34)
  • Risk of suicide during treatment with paroxetine vs no use; HR: 0.71 (95% CI 0.37–1.35)
  • Risk of suicide during treatment with citalopram vs no use; HR: 1.16 (95% CI 0.59–2.25)
  • Risk of suicide during treatment with sertraline vs no use; HR: 0.38 (95% CI 0.16–0.93); the risk of suicide for fluoxetine and fluvoxamine are not reported, because results were not robust
  • The HR of suicide death during exposure to SSRI vs nonexposure to any antidepressant was 0.64 (95% CI 0.38– 1.07), with the risk being lower during exposure to lower doses of SSRI: 0.41 (95% CI 0.17–0.96)
  • Women were at much lower risk of suicide death than men. HR: 0.14(95% CI 0.09–0.22)
  • Results for the subgroup of patients who had not received any antidepressant medication during 180 days prior to index date:
    SSRI vs nonexposure; HR: 0.72 (95% CI 0.39–1.34)
    Other antidepressants vs nonexposure; HR: 1.65 (95%CI 0.65–4.22)
    Both SSRI and another antidepressant vs nonexposure; HR: 2.01 (95%CI 0.46–8.75)
Adjusted Risk of poisoning events (Cox regression models with time-dependent exposure):
  • SSRI versus nonexposure to any antidepressant; HR: 1.16 (95%CI 1.07–1.25)
  • Risk of poisoning events during exposure to paroxetine; HR: 1.18 (95%CI 1.06–1.30)
  • Risk of poisoning events during exposure to citalopram; HR: 1.23 (95%CI 1.08–1.40)
  • Risk of poisoning events during exposure to sertraline; HR: 1.05 (95%CI 0.93–1.18)
  • Risk of poisoning events during exposure to fluvoxamine; HR: 1.45 (95%CI 1.23–1.71)
  • Risk of poisoning events during exposure to fluoxetine; HR: 0.93 (95%CI 0.74–1.16)
ITT: N.A
Post randomization exclusions: N.A
ATTRITION:Overall Attrition: N.A

Withdrawals due to adverse events: N.A
Withdrawals due to lack of efficacy: N.A
Differential Attrition: N.A
ADVERSE EVENTS:see main results
QUALITY RATING:Fair
STUDY:Authors: Rapaport ME, et. al.77
Year: 1996
Country: US
FUNDING:Solvay Pharmaceuticals, Upjohn
DESIGN:Study design: RCT
Setting: Multi-center (6 sites)
Sample size: 100
INTERVENTION:
Drug:FluvoxamineFluoxetine
Dose:100–150 mg/d20–80 mg/d
Duration:7 weeks7 weeks
INCLUSION:Male and female outpatients; 18–65 years; met DSM-III-R criteria for MDD; minimum HAM-D (21-item) score of 20; minimum score of 2 on the depressed mood item
EXCLUSION:Any primary DSM-IV Axis I disorder diagnosis other than MDD; acute suicidality; unstable medical conditions; history of seizure; had been treated with study medications; history of substance abuse or dependence; pregnancy and lack of appropriate birth control for women of child-bearing age
OTHER MEDICATIONS/ INTERVENTIONS:Chloral hydrate
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: fluoxetine: 38.6; fluvoxamine: 40.0
Gender (% female): fluoxetine: 63; fluvoxamine: 61
Ethnicity: 95% white; 5% other
Other population characteristics: NR
OUTCOME ASSESSMENT:Measures: HAM-D-21, HAM-A, CGI-S, Raskin–Covi Scale, Hopkins Symptom Checklist, TESS (Specific treatment- emergent signs and symptoms) Barnes Akathisia Scale, Modified Scale for Suicidal Ideation

Timing of assessments: Primary outcome measures weekly; secondary outcome measures at baseline and endpoint
RESULTS:
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes (7)
ATTRITION:Loss to follow-up: 11%
Withdrawals due to adverse events: 4%
Loss to follow-up differential high: No
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Schatzberg et al.82
Year: 2002
Country: US
FUNDING:Organon Pharma
DESIGN:Study design: RCT
Setting: Multi-center
Sample size: 255
INTERVENTION:(There was extension phase to 16 weeks but only included subjects who had favorable response during the first part of the study)
Drug:MirtazapineParoxetine
Dose:15–45 mg/d20–40 mg/d
Duration:8 weeks8 weeks
INCLUSION:Min. age of 65 years; DSM IV criteria for single or recurrent MDD; MMSE score > 25% for age and education; min. score of 18 on HAM-D17
EXCLUSION:HAMD decrease > 20% between screening and baseline; untreated or unstable clinically significant medical condition or lab/physical exam abnormality; history of seizures; recent drug or alcohol abuse or any principal psych condition other than MDD; presence of psychotic features; suicide attempt in current episode; use of MAOI within 2 weeks, or other psychotropics or herbal treatments within 1 week; use of paroxetine or mirtazpine for the current episode; ECT therapy within 6 months; use of treatment for memory deficits; prior intolerance or lack of efficacy to mirtazapine or paroxetine in the past; patients who failed more than one adequate trial of an antidepressant for the current episode
OTHER MEDICATIONS/ INTERVENTIONS:Chloral hydrate or zolpidem for sleep induction; therapy for conditions like DM, hypothyroidism, high blood pressure, chronic respiratory conditions was allowed if they had been receiving for at least 1 month prior to screening visit.
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 72
Gender (% female): Mirtazapine: 63%, paroxetine: 64%
Ethnicity: Not reported
Other population characteristics: Not reported
OUTCOME ASSESSMENT:Measures: HAM-D-17, CGI-S, CGI-I
Timing of assessments: Baseline, weeks 1, 2, 3, 4, 6, 8
RESULTS:
  • Mean Ham-D-17 scores significantly lower with mirtazapine at week 1, 2, 3, 6 but no difference at 8 week endpoint
  • Trend towards higher response and remission rates with mirtazapine but only significant difference at 2 weeks (response) and 6 weeks (remission)
  • Time to response: mirtazapine mean 26 days, paroxetine 40 days; p = −0.016 for Kaplan-Meier plot comparing the two
  • No difference in CGI Improvement response
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes
ATTRITION:Loss to follow-up: 26.8%
Withdrawals due to adverse events: 20.4%; mirtazapine 14.8%, paroxetine 26.2% (p < 0.05)
Loss to follow-up differential high: No
ADVERSE EVENTS:
QUALITY RATING:Fair
STUDY:Authors: Schneeweis et al. 226
Year: 2010
Country: Canada – British Columbia
FUNDING:NIMH
DESIGN:Study design: Retrospective cohort study
Setting: Healthcare utilization database
Sample size: 287,543 mean follow-up 0.49 person-years
INTERVENTION:
Drug:CitalopramFluoxetineFluvoxamineParoxetineSertralineVenlafaxinemirtazapine, nefazodone, and trazodone
Dose:NANANANANANANA
Duration:see abovesee abovesee abovesee abovesee abovesee abovesee above
Sample size:45 52222 207969074 78036 13535 73228 316
INCLUSION:British Columbia residents who had antidepressant therapy initiated and had a recorded diagnosis of depression.
EXCLUSION:Existing bipolar disorder
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 46
Gender (female %): 56
Ethnicity (Caucasian %): NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Suicide
Secondary Outcome Measures: Suicide attempts
Timing of assessments: various
RESULTS:
  • ANALYSIS:
ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events:
Withdrawals due to lack of efficacy:
Differential Attrition:
ADVERSE EVENTS: See results
QUALITY RATING:Good
STUDY:Authors: Schneeweis et al. 227
Year: 2010
Country: Canada – British Columbia
FUNDING:NIMH
DESIGN:Study design: Retrospective cohort study
Setting: Healthcare utilization database
Sample size: 20906 mean follow-up 0.49 person-years
INTERVENTION:
Drug:CitalopramFluoxetineFluvoxamineParoxetineSertralineVenlafaxinemirtazapine, nefazodone, and trazodone
Dose:NANANANANANANA
Duration:see abovesee abovesee abovesee abovesee abovesee abovesee above
Sample size:351829221068522134892197940
INCLUSION:British Columbia residents who had antidepressant therapy initiated and had a recorded diagnosis of depression.
EXCLUSION:Existing bipolar disorder
OTHER MEDICATIONS/ INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 15
Gender (female %): 63
Ethnicity (Caucasian %): NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Suicide – fluoxetine was reference group
Secondary Outcome Measures: Suicide attempts
Timing of assessments: various
RESULTS:
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS:See results
QUALITY RATING:Good
STUDY:Authors: Schneider LS et al.228 and Nelson JC et al.229
Year: 2003 and 2007
Country: USA
FUNDING:Pfizer
DESIGN:Study design: RCT
Setting: Multicenter
Sample size: 752
INTERVENTION:
Drug:SertralinePlacebo
Dose:50–100 mgNA
Duration:8 weeks8 weeks
Sample size:360368
INCLUSION:60 years of age and older with major depression, nonpsychotic, single episode and recurrent, with a duration of at least four weeks and a HAMD score > 18
EXCLUSION:Depressive disorder with psychotic features, dementia, organic mental disorder, or mental retardation; a score < 24 on the MMSE; any psychotic disorder or bipolar disorder; drug or alcohol abuse or dependence within the previous 6 months (except nicotine); a history of seizure disorder; previous nonresponse, known hypersensitivity, or contraindication to sertraline; participation in an investigational drug trial within 3 months; significant suicide risk, a need for ECT, additional psychotropic drugs, or hospitalization; regular, daily use of benzodiazepines within 3 weeks, antidepressants within 2 weeks, use MAOIs or fluoxetine within 5 weeks; depot antipsychotic drug within 6 months; initiation of individual or group psychotherapy within 3 months; and any clinically significant unstable medical disorder that might affect study participation
OTHER MEDICATIONS/ INTERVENTIONS:As-needed use of zolpidem, up to 10 mg/day, or temazepam, up to 30 mg/day, for sleep during the first 4 weeks; drugs used as anti-inflammatories or in rheumatic disease and gout (40%), antihypertensive drugs (27%), hormone replacement therapy (41% of women), drugs for of hyperlipidemia (14%), thyroid and antithyroid drugs (12%), ulcer- healing drugs (11%), ß-adrenergic antagonists (11%), drugs for diabetes (7%), hypnotics and sedatives (6%), bronchodilators (5%), and corticosteroids (4%). Overall, 87% took concomitant medication.
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: Sertraline 70.0 Placebo 69.6
Gender (female %): Sertraline 54 Placebo 58
Ethnicity: 93% caucasian
Other population characteristics: HAMD Sertraline 21.4 Placebo 21.4
OUTCOME ASSESSMENT:Primary Outcome Measures: Clinical response and suicide ideation
Secondary Outcome Measures: Hamilton scale subscales, Patient Global Impression, Quality of Life Enjoyment and Satisfaction Questionnaire, MMSE, and 36-Item Short-Form Health Survey subscales
Timing of assessments: Baseline and weekly
RESULTS:
ANALYSIS:ITT: Yes
Post randomization exclusions: 19
Loss to follow-up differential high: no
ATTRITION:SertralinePlacebo
Loss to follow-up:87 (23%)65 (17%)
Withdrawals due to adverse events:14%5%
Withdrawals due to lack of efficacy:1%3%
ADVERSE EVENTS:Diarrhea 19% vs. 7% P ≤ 0.05
Headache 17% vs. 13% P ≤ 0.05
Nausea 16% vs. 5% P ≤ 0.05
Somnolence 10% vs.4% P ≤ 0.05
Insomnia 9% vs. 6% P ≤ 0.05
Dry mouth 8% vs. 6%
Dizziness 8% vs. 7%
Tremor 6% vs. <1% P ≤ 0.05
Fatigue 5% vs. 1% P ≤ 0.05
QUALITY RATING:Fair
STUDY:Authors: Segraves, et al.85
Year: 2000
Country: US
FUNDING:Glaxo Wellcome Inc
DESIGN:Study design: RCT
Setting: Multi-center
Sample size: 248
INTERVENTION:
Drug:SertralineBupropion
Dose:50–200 mg/d100–300 mg/d
Duration:16 weeks16 weeks
INCLUSION:Received a DSM-IV diagnosis of moderate to severe depression with a minimum duration of 4 weeks and a maximum duration of 24 months; ≥ 18 years of age; in a stable relationship, have normal sexual functioning and sexual activity at least once every 2 weeks
EXCLUSION:Predisposition to seizure; pregnancy; alcohol or substance abuse; eating disorder; suicidal tendencies; prior treatment with bupropion or sertraline; used any psychoactive drug within 1 week of study
OTHER MEDICATIONS/INTERVENTIONS:None reported
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 39
Gender (% female): Sertraline: 48%, bupropion: 48%
Ethnicity: (% white) Sertraline: 94%, bupropion: 93%
Other population characteristics: No significant differences in diagnosis
OUTCOME ASSESSMENT:Measures: Sexual function assessment, Sexual desire disorder, Sexual arousal disorder, Orgasm dysfunction, Premature ejaculation, patient rated overall sexual satisfaction on 6 point Likert scale
Timing of assessments: Baseline, weeks 1, 2, 3, 4, 6, 8, 12, 16
RESULTS:

Significantly more sertraline patients developed a sexual dysfunction compared to bupropion patients; p < 0.001 for men and women p < 0.05 for sexual desire disorder

  • Overall sexual satisfaction (patient-rated) significantly more improved in bupropion treated patients. Men (p < 0.05) significant difference at day 21, 28, 42, and 56. Women (p < 0.01) beginning at day 56 and continuing to end
ANALYSIS:ITT: Yes
Post randomization exclusions: Yes
ATTRITION:Loss to follow-up: 31.5%; bupropion: 29%, sertraline: 34%
Withdrawals due to adverse events: 1.6%; bupropion 0%, sertraline 1.6%
Loss to follow-up differential high: Yes
ADVERSE EVENTS:Not reported
QUALITY RATING:Fair
STUDY:Authors: Strombom et al.230
Year: 2008
Country: USA
FUNDING:Eli Lilly
DESIGN:Study design: Data mining; FDA AERS database, claims database
Setting: Database
Sample size: 27,328
INTERVENTION:
Drug:DuloxetineVenlafaxine
Dose:NRNR
Duration:NRNR
Sample size:1366413664
INCLUSION:Patients taking duloxetine or venlafaxine
EXCLUSION:NA
OTHER MEDICATIONS/INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: NA
Mean age: NR
Gender (female %): NR
Ethnicity (Caucasian %): NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Hepatic events
Secondary Outcome Measures: none
Timing of assessments: NR
RESULTS:Similar rates of hepatic events between duloxetine compared with venlafaxine.
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS: See results
QUALITY RATING:NA
STUDY:Authors: Targownik et al. 231
Year: 2009
Country: Canada
FUNDING:Astra Zeneca Canada, Janssen-Ortho Canada
DESIGN:Study design: retrospective case-control study
Setting: outpatient
Sample size: 70,142
INTERVENTION: = exposureCases (diagnosis consistent with upper gastrointestinal bleeding, at least one overnight stay in hospital)Controls (had NOT been admitted to hospital for an upper gastrointestinal bleeding)
Drug:SSRIs(+venlafaxine) alone, PPI alone, NSAID alone, SSRI+PPI, SSRI+NSAID, NSAID+PPI, SSRI+NSAID+PPI)SSRIs(+venlafaxine) alone, PPI alone, NSAID alone, SSRI+PPI, SSRI+NSAID, NSAID+PPI, SSRI+NSAID+PPI)
Dose:NSNS
Duration:SSRI duration: <28days, 29–90days, >91daysSSRI duration: <28days, 29–90days, >91days
Sample size:1,55268,590
INCLUSION:Patient cohort consisted of all Manitobans over the age of 18 who had maintained continuous enrolment in the provincial health-care plan between 1 April 1995 and 31 March 2007.
Cases consisted of all subjects who were admitted to hospital (at least one overnight stay in hospital) with a most responsible diagnosis consistent with upper gastrointestinal bleeding (UGIB).
Among individuals who had UGIB secondary to an esophageal lesion, only non-variceal bleeds were included so as to exclude bleeds with different mechanistic etiologies.
EXCLUSION:Subjects who were admitted to hospital with UGIB before 1 October 1995 were excluded to ensure cases had at least of 180 days of prescription drug dispensation data available before the event date.
Controls excluded who were hospitalized for other indications on the day of the case’s event, such that all cases and controls were ambulatory in the community on the index date.
OTHER MEDICATIONS/INTERVENTIONS:Also tracked the use of other prescription medications believed to affect the risk of upper gastrointestinal complications, including warfarin, systemic corticosteroids, clopidogrel, and H2-receptor antagonists.
The drug database is unable to track the use of medications available without a prescription, such as aspirin. As such, a history of cardiovascular disease was considered to be a surrogate for aspirin use.
POPULATION CHARACTERISTICS:Groups similar at baseline: matched on age (within +/− 3 years), sex and overall medical comorbidity;
Mean age: cases: 68.9 yrs (± 15.2); controls: 69.8 ± (14.7)
Gender (male %): cases: 57.6%; controls: 56.4%
Cases are suffering from more comorbidities, especially cardiovascular diseases (32.2% vs. 16.8%) P<0.001.
OUTCOME ASSESSMENT:Primary Outcome Measures: gastro-intestinal bleeding
Secondary Outcome Measures: NA
Timing of assessments: NA
RESULTS:After adjusting for confounders, subjects who were admitted with UGIB were 1.43 times more likely to have been prescribed an SSRI (95 % CI 1.09 – 1.89) and were 3.17 times more likely to be using both an SSRI and NSAID (95 % CI: 2.01 – 5.00).
Subjects using both an SSRI and an NSAID were no more likely to develop UGIB than were users of an NSAID alone (OR, 1.21; 95 % CI, 0.76 – 1.93).
Recent past SSRI use was not associated with an increased risk of UGIB (OR, 0.99; 95 % CI, 0.63 – 1.56).
Overall, 4,371 average risk patients would have to receive an SSRI to promote the development of one additional case of non-variceal UGIB.
This study confirms that SSRI use is associated with a modest increase in the risk of UGIB, but did not detect an increase in the risk of UGIB events when an SSRI is combined with an NSAID, cotherapy with PPIs is able to significantly reduce the risk of SSRI-associated UGIB.
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS:see results
QUALITY RATING:Fair
STUDY:Authors: Tiihonen et al.232
Year: 2006
Country: Finland
FUNDING:EVO financing (special government subsidies) from Niuvanniemi Hospital.
DESIGN:Study design: Observational cohort
Setting: Nationwide
Sample size: 15,390
INTERVENTION:
Drug:Various
Dose:Various
Duration:Mean follow-up 3.4 years
Sample size:15390
INCLUSION:All individuals in Finland who were hospitalized with a diagnosis of suicide attempt from January 1, 1997, to December 31, 2003 (the first hospital treatment period was considered as the index period). and were at least 10 years old when the index hospitalization began.
EXCLUSION:Psychosis diagnosis
OTHER MEDICATIONS/INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: NA
Mean age: 38.8
Gender (female %): 51.5
Ethnicity: NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: relative risk (RR) of completed suicides, suicide attempts leading to hospitalization, and overall mortality during TCA (amitriptyline or doxepin hydrochloride), SSRI (fluoxetine, citalopram hydrobromide, paroxetine hydrochloride, sertraline, or fluvoxamine maleate), and SNA (mianserin hydrochloride, mirtazapine, or venlafaxine hydrochloride) treatment vs no antidepressant use
Secondary Outcome Measures: NA
Timing of assessments: various
RESULTS:Adjusted RR (95% CI)
ANALYSIS:ITT: NA
Post randomization exclusions: NA
Loss to follow-up: NA
ATTRITION:N/A
ADVERSE EVENTS:
  • See results
QUALITY RATING:Fair
STUDY:Authors: Tourian et al.233
Year: 2010
Country: Multinational
FUNDING:Wyeth and Phizer
DESIGN:Study design: Pooled analysis
Number of patients: 2950
AIMS OF REVIEW:To assess the risk of increased suicidal thoughts and behavior (suicidality) with desvenlafaxine (administered as desvenlafaxine succinate) in patients with major depressive disorder (MDD).
STUDIES INCLUDED IN REVIEW9 RCTs
TIME PERIOD COVERED:NR
CHARACTERISTICS OF INCLUDED STUDIES:RCTs, placebo controlled
CHARACTERISTICS OF INCLUDED POPULATIONS:Adult outpatients meeting DSM-4 criteria for MMD. Symptoms for at least 30 days
CHARACTERISTICS OF INTERVENTIONS:Placebo or desvenlafaxine 50 to 400 mg/d in 5 fixed-dose and 4 flexible-dose studies. Desvenlafaxine dose groups in the fixed- dose studies were 50 mg/d (included in 2 studies), 100 mg/d (3 studies), 200 mg/d (3 studies), and 400 mg/d (3 studies). In the flexible-dose studies, dose ranges were 100 to 200 mg/d (1 study) and 200 to 400 mg/d (3 studies). Two of the flexible-dose studies included venlafaxine extended-release treatment arms. The double-blind treatment period for all studies was 8 weeks, followed by a 1- to 2-week taper period
MAIN RESULTS:Completed suicide Single study OR 1.03 (0.04–25.56) P = 0.984
Suicide attempt; pooled analysis OR 0.95 (0.19–4.73) P = 0.952
Suicidal ideation; pooled analysis OR 0.77 (0.22–2.26) P = 0.677
Overall suicidality; pooled analysis OR 0.97 (0.31–2.84) P = 0.907
ADVERSE EVENTS:There were no significant differences between groups in the risk for any class of suicide-related events, including completed suicide or suicide attempt.
COMPREHENSIVE LITERATURE SEARCH STRATEGY:No
STANDARD METHOD OF APPRAISAL OF STUDIES:No
QUALITY RATING:N/A
STUDY:Authors: Trifiro, Dieleman, Sen, Gambassi, Sturkenboom 234
Year: 2010
Country: Netherlands
FUNDING:NR
DESIGN:Study design: case-control study (secondary data analysis)
Setting: community-dwelling elderly patients; primary care
Sample size: 492,272
INTERVENTION: = exposureCases (ischemic stroke)Controls
Drug:Current, past and non-use of any antidepressant (AD): tricyclic antidepressants, SSRIs, or other ADCurrent, past and non-use of any antidepressant (AD): tricyclic antidepressants, SSRIs, or other AD
Dose:Dose was also consideredDose was also considered
Duration:Cumulative number of prescription days during the follow-up period: short-term use (<=180 days) or long-term use (>= 181 days)Cumulative number of prescription days during the follow- up period: short-term use (<=180 days) or long-term use (>= 181 days)
Sample size:996491,276
INCLUSION:Persons 65 years and older registered in the Integrated Primary Care Information database (1996–2005). Cases were all patients with a validated first ischemic stroke. Controls were matched on year of birth, sex, and index date. Exposure to antidepressants was divided in current, past, and nonuse and further categorized by type (SSRI, tricyclic, and other antidepressants), dose, and duration.
EXCLUSION:Patients who had a recorded diagnosis of TIA or stroke in the medical history before the study entry or patients with a diagnosis of cerebral tumor either before or during the study period.
OTHER MEDICATIONS/INTERVENTIONS:Prior use of cardiovascular medications and concomitant use of psychotropic drugs or other drugs (systemic corticosteroids, antibiotics, NSAIDs) were considered as covariates
POPULATION CHARACTERISTICS:Cases and Control Groups similar: yes
Age groups:cases:controls
 65–74 years:32.2%47.4%
 75–84 years:44.9%44.9%
 >= 85 years:22.9%7.7%
Gender (female %): cases: 58.2%; controls: 61.8% (matching)
Ethnicity (Caucasian %): NR
Other population characteristics: No relevant differences in smoking habits, cardiovascular diseases, and other diseases potentially related to stroke or neuropsychiatric diseases.
OUTCOME ASSESSMENT:Primary Outcome Measures: risk of ischemic stroke between users of antidepressants and nonusers
Secondary Outcome Measures: NA
Timing of assessments: NA (secondary analysis; observation period: January 1, 1996 – December 31, 2005)
RESULTS:
  • Current use of SSRIs was associated with an increased risk of ischemic stroke compared with non-use; OR: 1.55 (95% CI 1.07–2.25), whereas no significant associations were found for current use of TCA OR: 1.18 (95% CI 0.73–1.91) or other antidepressants OR: 1.01 (95% CI 0.45–2.25).
  • There was no dose effect observed on the risk of ischemic stroke for current users of any antidepressant type.
  • A duration effect was observed: Shorter use (<=180 days) of SSRIs was associated with a larger risk increase OR: 2.07 (95% CI 1.24–3.46) than longer use (>180 days) OR: 1.14 (95% CI 0.65–1.97).
  • For patients with depression as an indication for treatment, the risk of ischemic stroke with SSRIs use OR: 1.99 (95% CI 1.20–3.30) was higher than that with TCAs OR: 1.07 (95% CI 0.43–2.65), although the difference was not statistically significant.
ANALYSIS:ITT: NA
Post randomization exclusions: NA
ATTRITION:Overall Attrition: NA
Withdrawals due to adverse events: NA
Withdrawals due to lack of efficacy: NA
Differential Attrition: NA
ADVERSE EVENTS:See results
QUALITY RATING:Good
STUDY:Authors: Valuck R et al.235
Year: 2004
Country: USA
FUNDING:Unfunded
DESIGN:Study design: Retrospective cohort
Setting: Health Insurance database
Sample size: 24119
INTERVENTION:
Drug:SSRIs-citalopram escitalopram fluoxetine fluvoxamine paroxetine, sertraline venlafaxineOthers- Bupropion mirtazapine nefazadone trazodoneNoneMultiple
Dose:VariousVariousVariousVarious
Duration:Mean 1.36 yearsMean 1.36 yearsMean 1.36 yearsMean 1.36 yrs
Sample size:459549217313173131674
INCLUSION:adolescents 12–18 years who received either a diagnosis of MDD or an antidepressant medication (or both) between January 1998 and March 2003. A retrospective cohort was created for adolescents with new starts of depression treatment
EXCLUSION:Previous depression claims, antidepressant use or psychotherapy
OTHER MEDICATIONS/INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 12–6.3%, 13–8.7%, 14–11.8%, 15–16.0%, 16–19.8%, 17–20.6%, 18–16.%
Gender (female %): 63
Ethnicity: NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Suicide attempt
Secondary Outcome Measures:
Timing of assessments: Various
RESULTS:
  • Crude rates of Suicide attempt rate per person- month of follow-up (%) SSRI 0.13 Other 0.11 Multiple 0.11 None 0.07 Total 0.09
  • Results from cox proportionate model shows that the hazard ratios (95% CI) for SSRI 1.59 (0.89 to 2.82) P = 0.116, Other 1.03 (0.43 to 2.42), Multiple 1.43 (0.70 to 2.89) P= 0.325, None 1.00 referent.
  • Other variables of interest include, female 1.97 (1.38 to 2.83) P < 0.001, duration of use >180 days 0.34 (0.21 to 0.55) P < 0.001
ANALYSIS:ITT: NA
Post randomization exclusions: NA
Loss to follow-up: NA
ATTRITION:NA
Withdrawals due to adverse events:
Withdrawals due to lack of efficacy:
Loss to follow-up differential high:
ADVERSE EVENTS: See results
QUALITY RATING:Fair
STUDY:Authors: Vanderburg et al.236
Year: 2009
Country: NA
FUNDING:Pfizer
DESIGN:Study design: Pooled analysis (of Pfizer-sponsored RCTs of sertraline)
Number of patients: 19,923
AIMS OF REVIEW:To identify possible suicide-related events in completed placebo-controlled Pfizer-sponsored trials of sertraline in adult patients with various psychiatric indications and to assess risk of suicidality with sertraline vs. placebo
STUDIES INCLUDED IN REVIEW126 placebo-controlled, double-blinded RCTs
TIME PERIOD COVERED:Mid-1980s – mid-2000s
CHARACTERISTICS OF INCLUDED STUDIES:126 RCTs (Pfizer-sponsored completed studies, placebo-controlled, double-blind, all-duration, all indication, Phases 2–4), conducted between mid-1980s – mid-2000s; also including studies with no suicidal event or studies with < 20 subjects per treatment arm; relapse prevention studies were included
CHARACTERISTICS OF INCLUDED POPULATIONS:Adults, both sexes; various psychiatric indications (MDD and non-MDD, including bipolar disorder, generalized anxiety disorder, panic disorder, obsessive-compulsive disorder, substance abuse, dysthymia, atypical depression, Post-traumatic Stress Disorder, generalized social phobia, bulimia nervosa, premenstrual dysphoric disorder) as well as some non-psychiatric indications (obesity, smoking cessation, fibromyalgia); during randomized phase of RCT or within 1 day after stopping randomized treatment
CHARACTERISTICS OF INTERVENTIONS:sertraline vs. placebo (all duration, including short-term trial duration < 17 weeks) during randomized phase of RCT or within 1 day after stopping randomized treatment
MAIN RESULTS:Suicidality was the primary outcome and classified using the Columbia Classification Algorithm of Suicide Assessment:
1. Completed Suicide, 2. Suicide Attempt, 3. Preparatory Actions Towards Imminent Suicidal Behavior, 4. Suicidal Ideation;
  • 99 suicidality events were identified among 19,923 sertraline- and placebo-treated subjects; 4 cases of completed suicides among 10,917 sertraline-treated subjects with an incidence of 0.04% (95% CI = 0.01 to 0.09) and 3 cases among 9,006 placebo treated subjects with an incidence of 0.03% (95% CI = 0.01 to 0.10).
  • No statistically significant differences between sertraline and placebo group in any of the individual suicidality groups or in all groups combined (i.e. short-term studies vs. all-duration studies; MDD vs. non-MDD-indication studies; age groups (< 25 years; 25–64 years, ≥ 65 years)
  • In all-duration psychiatric studies the RR of suicidality combined was 0.96, 95% CI (0.64 – 1.44); in the all-duration psychiatric studies age group analyses the RR of suicidality combined was 0.60, 95% CI (0.16 – 2.23) for those aged < 25 years, and 0.88, 95% CI (0.55 – 1.39) in the age group 25 – 64 years and 1.32, 95% CI (0.32 – 5.52) in the age group ≥ 65 years;
ADVERSE EVENTS:see results for detail
COMPREHENSIVE LITERATURE SEARCH STRATEGY:No (see comments)
STANDARD METHOD OF APPRAISAL OF STUDIES:No (see comments)
QUALITY RATING:N/A
STUDY:Authors: Vanderkooy et al.237
Year: 2002
Country: Canada
FUNDING:NR
DESIGN:Study design: Prospective Observational
Setting: Tertiary care clinic
Sample size: 193
INTERVENTION:
Drug:VenlafaxineParoxetineSertralineMoclobemideBuproppion
Dose:NRNRNRNRNR
Duration:8 weeks8 weeks8 weeks8 weeks8 weeks
Sample size:6255372415
INCLUSION:Patients that completed 8 weeks of treatment for depression
EXCLUSION:NA
OTHER MEDICATIONS/INTERVENTIONS:NR
POPULATION CHARACTERISTICS:Groups similar at baseline: Yes
Mean age: 39.5
Gender (female %): 62%
Ethnicity: NR
Other population characteristics:
OUTCOME ASSESSMENT:Primary Outcome Measures: Remission and adverse events
Timing of assessments: Baseline and 6 weeks
RESULTS:
ANALYSIS:ITT: No
Post randomization exclusions: NA but 24 or 11% noncompleters
ATTRITION:Loss to follow-up: bupropion 12%, moclobemide 16%, paroxetine 23%, sertraline 24%, venlafaxine 13%
Withdrawals due to adverse events: NR
Withdrawals due to lack of efficacy: NR
Loss to follow-up differential high: No
ADVERSE EVENTS:Adverse events %
Venlafaxine vs. paroxetine vs. sertraline
Nervousness 11 vs. 9.1 vs. 16
Agitation 18 vs. 11 vs. 19
Tremor 11 vs. 3.6* vs. 16
Myoclonus 9.7 vs.13 vs.14
Fatigue 24 vs. 13 vs. 22
Dizziness 9.7 vs. 11 vs. 14 8
Postural hypotension 15 vs. 7.3* vs. 22
Somnolence 27 vs. 29 vs. 32
Increased sleep 6.5 vs. 7.3 vs. 14
Decreased sleep 26 vs. 13 vs. 14
Sweating 27 vs. 27 vs. 32
Flushing 11 vs. 13 vs. 14
Edema 1.6 vs. 1.8 vs. 8.1
Headache 26 vs. 18 vs. 22
Blurred vision 9.7 vs. 15 vs. 14
QUALITY RATING:Fair
STUDY:Authors: Vestergaard et al. 238
Year: 2008
Country: Denmark
FUNDING:
DESIGN:Study design: case-control study (secondary data analysis, data linkage)
Setting: in/-outpatient, community (population-based)
Sample size: 498,617
INTERVENTION: = exposureCasesControls
Drug:Exposure to any antidepressant (AD): tricyclic antidepressants, SSRIs, SNRIs, MAOIs, NaSSA, tetracyclic antidepressantsExposure to any antidepressant: tricyclic antidepressants, SSRIs, SNRIs, MAOIs, NaSSA, tetracyclic antidepressants)
Dose:NRNR
Duration:0–5 years (yrs)0–5 years
Sample size:124,655373,962
INCLUSION:Civil registry number (entry in civil registration system, coverage in hospital discharge register a/o pharmacy database)
EXCLUSION:Criteria NR (various sensitivity analyses for confounder analysis)
OTHER MEDICATIONS/INTERVENTIONS:Corticosteroids, antiepileptics, neuroleptics, anxiolytics, hypnotics, sedatives, lithium (adjustments for ever use were made)
POPULATION CHARACTERISTICS:Groups similar at baseline: only for age and sex
Mean age: cases: 43.44 yrs ± 27.39; controls: 43.44 yrs ± 27.39
Gender (female %): cases: 51,8%; controls: 51,8% (matching)
Ethnicity (Caucasian %): NR
Other population characteristics: relevant differences for marital status, comorbidity, income, previous fracture
OUTCOME ASSESSMENT:Primary Outcome Measures: conditional OR of fracture (any, hip, spine, forearm) of single AD (multiple adjustments; analyses done for average daily dose; cumulated use, duration of use)
Secondary Outcome Measures: NA
Timing of assessments: NA (secondary analysis; data on AD use available for 0–5 years), dosages were calculated as DDD (defined daily dose) per day
RESULTS:Only statistically significant associations of relevant drugs shown here: dose-response relationship for several SSRIs (citalopram p<0.01, fluoxetine p< 0.03, fluvoxamine p< 0.74, paroxetine p<0.078, sertraline p<0.01); risk of any fracture (all adjusted ORs, dosage categories <=0,25 DDD; 0,25–0,5 DDD; >= 0,5 DDD); from lowest to highest dose: citalopram 1.11 (1.06–1.16); 1.31 (1.21–1.41); 1.38 (1.33–1.44); fluoxetine 1.06 (1.00–1.13); 1.16 (1.01–1.33); 1.20 (1.09–1.32); paroxetine (only at highest dose) 1.21 (1.10–1.33); sertraline (only at highest dose) 1.25 (1.16–1.34); mirtazapine (only at medium dose) 1.22 (1.05–1.41); In general, the increase in ‘relative risk’ of hip fractures was larger than the increase in other fracture types; risk for all fracture types statistically significantly increased for citalopram, at highest dose also for sertraline; greatest risk increase (adj OR) for hip fracture 1,98 (1,82–2,16) for citalopram; statistically significant associations for SSRIs and duration of use with a tendency to decline with time (duration of use. categorized <=0,5 yrs; 0,5–1 yrs; 1,1–2,5 yrs; >=2,5 yrs); statistically significant for citalopram, fluoxetine,