WHO 1998.

Methods	Randomised, double‐blind, multinational trial. Random number generation done centrally. Double‐blinded by use of identical placebos. Allocation concealment achieved by sealed, sequentially‐numbered, tinted bottles, filled and labelled by the manufacturer
Participants	1998 healthy women at 21 centres worldwide Included women with regular menstrual periods, aged 18‐45 years, who had attended the clinic within 72 h of a single act of unprotected intercourse Excluded women who were breastfeeding, on hormonal contraception in the current cycle and those with uncertain dates of last menstrual period 1955 women into the final analysis
Interventions	Yuzpe (ethinyl oestradiol 100 μg + LNG 0.50 mg, repeated after 12 h) vs LNG 0.75 mg, twice, 12 h apart
Outcomes	Observed number of pregnancies, side effects and changes in menstrual pattern
Notes	Loss to follow‐up: Yuzpe 18/997 (1.8%); LNG 25/1001 (2.5%) Post‐randomisation exclusion (ITT analysis) not reported Observed pregnancy/expected pregnancy/total number of women: Yuzpe 31/72/979; LNG 11/75.3/976
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"The unit of randomisation was the individual woman. We used a computer generated randomisation sequence developed in Geneva to assign participants to treatment groups. Each centre received assignments by random permuted blocks with a fixed block size of ten."
Allocation concealment (selection bias)	Low risk	"The allocation was concealed by use of sealed, sequentially numbered, tinted pill bottles, which were filled and labelled by the manufacturer." "The allocation sequence was kept in Geneva, and assignments were not revealed to investigators or participants during the trial."
Blinding (performance bias and detection bias) All outcomes	Low risk	"Clinicians and participants were unaware of the next assignment." "Double blinding was maintained throughout the trial. Each pill bottle contained two identical tablets. Bottles containing a levonorgestrel tablet had an identical placebo tablet. The supplier formulated, especially for the trial, tablets containing the Yuzpe regimen, of identical appearance to the levonorgestrel tablets."
Incomplete outcome data (attrition bias) All outcomes	Low risk	"We analysed the data by intention to treat." Also explained loss to follow‐up reasons
Selective reporting (reporting bias)	Low risk	Reported planned primary and secondary outcomes
Other bias	Low risk	None detected