Table S1. Characteristics of Included Studies

Table S1. Characteristics of included studies

Buvat 2009

Methods / 130 centers worldwide, December 2004 to October 2006
24 weeks with 1 week withdrawal phase
Participants / DSM-IV-TR criteria, PE for ≥6 mo, indicated at least moderate PE-related distress or interpersonal difficulty, and reported an IELT of ≤2 min in ≥75% of evaluable events during a 4-wk screening
baseline IELT: 0.9min
excluded:
· history of medical or psychiatric illness
· medical events associated with the onset of PE
· other sexual dysfunction (men and partner)
· uncontrolled hypertension or cardiac impairment
· hypersensitivity to SSRIs
· concomitant use of SSRIs

alcohol abuses

Interventions / dapoxetine 30mg or dapoxetine 60 mg or placebo
Outcomes / primary: IELT
Notes / serious AEs occurred in 1.0%, 0.8%, and 1.0% of subjects treated with dapoxetine 30 mg and dapoxetine 60 mg and placebo
we extracted results from 30mg dose
conclusion:
“dapoxetine significantly improved all aspects of PE and was generally well tolerated in this broad population.”

Risk of bias table

Bias / Authors' judgement / Support for judgement
Random sequence generation (selection bias) / Low risk / computer-generated randomization schedule
Allocation concealment (selection bias) / Low risk / assigned and coded using an interactive voice response system
Blinding of participants and personnel (performance bias) / Unclear risk / ns
Blinding of outcome assessment (detection bias) / Unclear risk / ns
Incomplete outcome data (attrition bias) / High risk / LOCF
1162 randomized:
385 placebo, 388 dapoxetine 30mg, 389 dapoxetine 60mg
546 (47%) dropouts
Selective reporting (reporting bias) / Low risk / all outcomes were reported
Other bias / High risk / conflict of interest

C-2002-012

Methods / 121 centers, US, June 2003 - June 2004
12 weeks
Participants / IELT, as measured by a partner-held stopwatch, of ≤ 2 minutes in at least three of four (or 75%) evaluable events occurring during a 2-week baseline period
baseline IELT: 0.91min
excluded:
· other sexual dysfunctions
· history of surgery or injury to the pelvis or spinal cord, chronic inflammation of the prostate or urethra
· concomitant use of SSRIs or tricyclic antidepressants
· history of major psychiatric disorder
· other forms of therapy for premature ejaculation (pharmacological or behavioural)

partners with self-reported female sexual dysfunction

Interventions / placebo, 30 mg dapoxetine, 60 mg dapoxetine
Outcomes / primary: IELT
Notes / we extracted results for dapoxetine 30mg
we calculated AEs from percentages reported in pooled analysis
comment:
EMA reference has no results for placebo, pooled analysis (Pryor et al) reports placebo IELT 1.75min (SD 2.21, n = 870)
from pooled analysis:
· any sexual AE: 4.5% dapoxetine, 1.9% placebo, RR 2.28 [1.30, 4.01]
· SEAs: 25 placebo, 20 dapoxetine
conclusion:
“On-demand dapoxetine is an effective and generally well tolerated treatment for men with moderate-to-severe premature ejaculation.”

Risk of bias table

Bias / Authors' judgement / Support for judgement
Random sequence generation (selection bias) / Low risk / stratified on the basis of average IELT per study center
Allocation concealment (selection bias) / Low risk / computerised interactive voice recognition system
Blinding of participants and personnel (performance bias) / Unclear risk / tablets in all groups were identical in appearance
comment:
not clear what aspects were matching
Blinding of outcome assessment (detection bias) / Unclear risk / ns
Incomplete outcome data (attrition bias) / High risk / LOCF
1294 randomized:
440 placebo, 429 dapoxetine 30 mg, 425 dapoxetine 60 mg
324 (25%) dropouts (percentage from pooled analysis)
Selective reporting (reporting bias) / High risk / from pooled analysis:
additional analysis was done, in which the baseline assessment was carried forward for patients without post-baseline assessments. -> results not reported
both studies were first analyzed separately and shown to have highly significant and similar results for the primary endpoint and all secondary endpoints.
comment:
no separate publicized study results given in, only pooled results. EMA reference has separate results
Other bias / High risk / conflict of interest

C-2002-013

Methods / 121 centers, US, June 2003 - June 2004
12 weeks
Participants / IELT, as measured by a partner-held stopwatch, of ≤ 2 minutes in at least three of four (or 75%) evaluable events occurring during a 2-week baseline period
baseline IETLT: 0.91min
excluded:
· other sexual dysfunctions
· history of surgery or injury to the pelvis or spinal cord, chronic inflammation of the prostate or urethra
· concomitant use of SSRIs or tricyclic antidepressants
· allergy to SSRIs
· history of major psychiatric disorder
· other forms of therapy for premature ejaculation (pharmacological or behavioural)
· partners with self-reported female sexual dysfunction

alcohol abuse

Interventions / placebo, 30 mg dapoxetine, 60 mg dapoxetine
Outcomes / primary: IELT
Notes / we extracted results from dapoxetine 30mg
we calculated AEs from percentages reported in pooled analysis
comment:
EMA reference has no results for placebo, pooled analysis reports placebo IELT 1.75min (SD 2.21, n = 870)
from pooled analysis:
· any sexual AE: 4.5% dapoxetine, 1.9% placebo, RR 2.28 [1.30, 4.01]
· SEAs: 25 placebo, 20 dapoxetine
conclusion:
“On-demand dapoxetine is an effective and generally well tolerated treatment for men with moderate-to-severe premature ejaculation.”

Risk of bias table

Bias / Authors' judgement / Support for judgement
Random sequence generation (selection bias) / Low risk / stratified on the basis of average IELT per study center
Allocation concealment (selection bias) / Low risk / computerised interactive voice recognition system
Blinding of participants and personnel (performance bias) / Unclear risk / tablets in all groups were identical in appearance
comment:
not clear what aspects were matching
Blinding of outcome assessment (detection bias) / Unclear risk / ns
Incomplete outcome data (attrition bias) / High risk / LOCF
1320 randomized:
430 placebo, 445 dapoxetine 30 mg, 445 dapoxetine 60mg
330 (25%) dropouts (percentage from pooled analysis)
Selective reporting (reporting bias) / High risk / from pooled analysis:
additional analysis was done, in which the baseline assessment was carried forward for patients without post-baseline assessments. -> results not reported
comment:
this ITT analysis would be very informative
both studies were first analyzed separately and shown to have highly significant and similar results for the primary endpoint and all secondary endpoints.
comment:
no separate publicized study results given in, only pooled results. EMA reference has separate results
Other bias / High risk / conflict of interest

Hellstrom 2004

Methods / 15 centers, US
6 weeks: 3 period (2w) crossover, 4 days washouts
Participants / IELT <2 min
baseline IELT: 1.01 min
excluded: ns
Interventions / placebo, dapoxetine 60mg on-demand (prn), dapoxetine 100mg
Outcomes / primary: IELT
Notes / found abstract only
we extracted 60mg dose
conclusion:
“dapoxetine HCl 60 mg and 100 mg prn appeared to be effective in improving PE as assessed by IELT.”

Risk of bias table

Kaufman 2009

Methods / 91 centers, US, Canada, November 2004 - August 2005
9 weeks with 1 week withdrawal assessment and 2 weeks follow-up
Participants / DSM-IV-TR criteria for PE, to have had PE for ≥6 months and to have reported at least ‘moderate’ distress or interpersonal difficulty related to their PE at baseline
Baseline IELT: ns
Excluded:
· a previous event or condition associated with PE (such as spinal trauma or pelvic surgery)
· another sexual dysfunction
· condition that affected overall physical or mental health status
· allergy to SSRIs
· history of drug abuse within the past 2 years

alcohol consumption more than two drinks per day

Interventions / placebo, dapoxetine 60mg as needed, 60mg daily
Outcomes / primary: withdrawal effects after abrupt discontinuation of dapoxetine in the once-daily arm
Comment: withdrawal effects were not reported in full report, but were reported in a conference abstract
unclear wether IELT was measured
Notes / we extracted 60mg as needed data
SAEs:
7 daily dapoxetine/ 2 dapoxetine as needed / 1 placebo
conclusion:
“dapoxetine reduced the personal distress and interpersonal difficulty associated with PE, and was associated with patient- reported improvements in their condition. The composite PRO was substantially greater than with placebo, as were all outcome measures.”

Risk of bias table

Bias / Authors' judgement / Support for judgement
Random sequence generation (selection bias) / Unclear risk / ns
Allocation concealment (selection bias) / Unclear risk / ns
Blinding of participants and personnel (performance bias) / Unclear risk / ns
Blinding of outcome assessment (detection bias) / Unclear risk / ns
Incomplete outcome data (attrition bias) / High risk / LOCF
1238 randomized:
245 placebo, 491 dapoxetine as needed, 502 dapoxetine daily
421 (34%) dropouts
comment:
ClinicalTrials.gov: 199 were randomized?
Selective reporting (reporting bias) / High risk / we extracted dapoxetine as needed data
results dapoxetine daily were not reported
many subgroup analyses, not clear wether these were prespecified
Other bias / High risk / conflict of interest

McMahon 2010

Methods / 52 centers, Asia, Pacific, March 2005 - June 2006
12 weeks
Participants / PE at least 6 months prior to enrollment and had a baseline IELT of 2 minutes or less in at least 75% of a minimum of four evaluable sexual intercourse events during a treatment- free 4-week baseline period, and reported at least “moderate” ejaculation-related personal distress or interpersonal difficulty.
baseline IELT: 1min
excluded:
· medical events that were associated with the onset of PE
· other sexual dysfunction
· medical or psychiatric illness
· uncontrolled hypertension
· hyperprolactinemia
· untreated hypothyroidism
· history of human immunodeficiency virus, hepatitis B surface antigen or hepatitis C
· concomitant use of SSRIs or tricyclic antidepressants

hypersensitivity to SSRIs

Interventions / placebo, dapoxetine 30 mg, dapoxetine 60 mg on demand
Outcomes / primary: IELT
Notes / we extracted dapoxetine 30mg dose
Serious treatment emergent AEs were reported by three subjects each in the placebo and dapoxetine 30 mg groups
conclusion:
“dapoxetine treatment significantly prolonged IELT and improved PEP measures and was generally well tolerated in men with PE in the Asia-Pacific region.”

Risk of bias table

Bias / Authors' judgement / Support for judgement
Random sequence generation (selection bias) / Unclear risk / ns
Allocation concealment (selection bias) / Unclear risk / ns
Blinding of participants and personnel (performance bias) / Unclear risk / ns
Blinding of outcome assessment (detection bias) / Unclear risk / ns
Incomplete outcome data (attrition bias) / High risk / LOCF
1067 randomized:
357 placebo, 354 dapoxetine 30mg, 356 dapoxetine 60mg
213 (20%) dropped out
Selective reporting (reporting bias) / Unclear risk / not sufficient information to judge
Other bias / High risk / conflict of interest

McMahon 2013

Methods / 69 centers worldwide, April 2010 - August 2011
12 weeks
Participants / Both with PE and ED
IELT of ≤ 2 minutes in a minimum of three out of four evaluable events, or in ≥75% of evaluable events if more than four evaluable events
International Index of Erectile Function-erectile function domain (IIEF-EF) score ≤ 21
baseline IELT: 1.1min
excluded:
· previous event or condition associated with PE (e.g., spinal trauma), presence of another sexual dysfunction in the man or his partner
· serious condition that affected overall physical or mental health status
· allergy to SSRIs
· history of drug abuse within the past 2 years

major psychiatric disorders

Interventions / placebo, dapoxetine flexible dose (30mg or 60mg)
Outcomes / primary: IELT
Notes / SEAs:
four subjects (1.6%) in the placebo group and three subjects (1.2%) in the dapoxetine group
1 death on dapoxetine (considered by the investigator to be not related to study drug)
conclusion:
“In men with PE and comorbid ED on a stable regimen of PDE5 inhibitor, dapoxetine provided meaningful treatment benefit and was generally well tolerated.”

Risk of bias table

Bias / Authors' judgement / Support for judgement
Random sequence generation (selection bias) / Low risk / Randomization was stratified by type of PDE5 inhibitor, baseline average IELT, region
computer-generated random sequence
Allocation concealment (selection bias) / Low risk / interactive voice response system
Blinding of participants and personnel (performance bias) / Unclear risk / matching placebo
comment
not clear what characteristics were matching
Blinding of outcome assessment (detection bias) / Unclear risk / ns
Incomplete outcome data (attrition bias) / High risk / LOCF
495 randomized:
245 placebo, 250 dapoxetine
66 (13%) dropouts
Selective reporting (reporting bias) / Unclear risk / an independent external statistics reporting group (SRG) conducted did one interim analysis during course of study when 268 subjects had completed 12 weeks of treatment
statistician was unblinded to the treatment assignments at the interim analysis
comment:
possible risk that study was unblinded and stopped early when results were significant or study prolonged to attain significant results
Other bias / High risk / conflict of interest
post-treatment follow up via telephone, 2 weeks after the end of treatment or early withdrawal visit, to collect information on AEs
comment:
posthoc collection of AEs

Safarinejad 2008

Methods / 1 center, Iran, February 2004 to March 2006
12 weeks
Participants / IELT of ≤ 2 min that occurred in more than 90% of intercourses
baseline IELT: 0.48min
excluded:
· organic cause of PE
· other sexual dysfunction
· neurological disorder
· depression
· psychiatric, or physical illness
· alcohol, drug, or substance abuse
· organic illness causing limitation in SSRI use
· use of psychotropic and antidepressant medication

serious relationship problems

Interventions / placebo, 30 mg dapoxetine twice a day (every 12h)
Outcomes / primary: IELT
Notes / all patients relapsed 3 months after treatment cessation
baseline IELT was 0.48min, this population had twice more severe PE than other studies
conclusion:
“dapoxetine has moderately better results in terms of IELT and intercourse satisfaction vs placebo without long-term benefit for the patient after it is withdrawn. Further studies are necessary to draw final conclusions on the efficacy of this drug in PE.”

Risk of bias table

Bias / Authors' judgement / Support for judgement
Random sequence generation (selection bias) / Low risk / randomization table generated by the method of random permuted blocks
Allocation concealment (selection bias) / Low risk / interactive voice response system
randomized sequence remained unknown to the patient and to the physician
a geographically and operationally independent from the study investigator did the randomization of the study
Blinding of participants and personnel (performance bias) / Low risk / placebo was a starch compound with the same color and size of dapoxetine
comment:
placebo seems matched but caveat! participants were informed on AEs during informed consent, potentially impairing blinding due to participants receiving placebo and experiencing few AEs
Blinding of outcome assessment (detection bias) / Low risk / the random sequence generation AND the allocation concealment AND the blinding of the participant had a low ROB and the assisting outcome assessor was NOT the caregiver that gave the intervention
Incomplete outcome data (attrition bias) / High risk / modified intention to treat analysis
comment:
dropouts were excluded from the analysis
212 randomized
23 (11%) dropouts
Selective reporting (reporting bias) / Low risk / all outcomes were reported
Other bias / Low risk / none

Footnotes

IELT: intravaginal latency time