McMahon 2013.

Study characteristics
Methods	Study design: randomized controlled, parallel‐group trial Setting/country: Berry Road Medical Center, the Leonards, Australia Dates study conducted: April 2010 to August 2011
Participants	Inclusion criteria: Age ≥ 18 years In a stable, monogamous heterosexual relationship for ≥ 6 months Planning to maintain this relationship for the duration of the study PE according to DSM‐IV‐TR criteria for PE for ≥ 6 months prior to enrolment and had a baseline IELT ≤ 2 minutes in ≥ 75% of a minimum of 4 evaluable sexual intercourse events during a treatment‐free 4‐week baseline period, and reported at least "moderate" ejaculation‐related personal distress or interpersonal difficulty Exclusion criteria: History of medical or psychiatric illness, including uncontrolled hypertension, hyperprolactinemia or untreated hypothyroidism History of medical events that were associated with the onset of PE Sexual dysfunction in either partner except PE in the man (including a history of ED based on IIEF Erectile Function domain score < 21 at screening) History of HIV, HBsAg or hepatitis C (except for people from Korea, inactive HBsAg carriers with normal liver function tests were allowed into the trial) Concomitant use of SSRIs or tricyclic antidepressants Known hypersensitivity to SSRIs or serotonin‐norepinephrine reuptake inhibitors Total number of participants randomized: 495 Total length of study: 18 weeks Group 1 (dapoxetine): Number of participants randomized: 250 Age (mean): 49.5 (SD 11.23) years Baseline IELT: NR Number of participants with primary PE: 92 (42.2%) Group 2 (placebo): Number of participants randomized: 245 Age (mean): 47.9 (11.96) years Baseline IELT: NR Number of participants with primary PE: 96 (45.9%)
Interventions	Group 1: dapoxetine 30 mg on‐demand, from week 4 up to 60 mg if tolerated + PDE5 inhibitor taken 1–3 hours prior to sexual intercourse Group 2: placebo daily + PDE5 inhibitor taken 1–3 hours prior to intercourse
Outcomes	Primary outcomes: IELT How measured: by female partner using a stopwatch Time points measured: 0, 4, 8, 12 weeks Secondary outcomes: Participant‐reported outcome measures How measured: CGIC in PE and PEP questionnaires Time points measured: every 4 weeks Safety outcomes: Adverse effects How measured: reported by participants Time points measured: anytime
Funding sources	Janssen Research & Development, LLC funded this study (R096769PRE3008) and provided formal review of the article.
Declarations of interest	Janssen Research & Development, LLC funded this study (R096769PRE3008) and provided formal review of the article. Bradford Challis, PhD, an employee of the company, provided writing assistance for the manuscript.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Randomization relied on a computer‐generated random sequence and an interactive voice response system."
Allocation concealment (selection bias)	Low risk	Quote: "Randomization relied on a computer‐generated random sequence and an interactive voice response system."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Only the IDMC statistician was unblinded to the treatment assignments at the interim analysis." Therefore, can assume that participants, investigators and personnel were blinded.
Blinding of outcome assessment (detection bias) Participant‐reported outcomes	Low risk	Likely appropriately blinded. Quote: "subjects were instructed to administer study drug (dapoxetine or matching placebo)."
Blinding of outcome assessment (detection bias) Investigator‐assessed outcomes	Low risk	Quote: "Only the IDMC statistician was unblinded to the treatment assignments at the interim analysis." Therefore, can assume that investigators and personnel were blinded.
Blinding of outcome assessment (detection bias) IELT	Low risk	Objective measurement that was unlikely to be influenced by blinding.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Large number of discontinuations and not balanced between groups (29/250 in dapoxetine arm and 37/245 in placebo arm).
Selective reporting (reporting bias)	Low risk	As reported in protocol in ClinicalTrials.gov.
Other bias	Low risk	No additional sources of bias identified.