Brazil 2016.

Methods	Allocation concealment: sealed and numbered opaque envelopes. Two arms
Participants	100 singleton pregnant women with PE, defined as development of HT after 20 weeks of gestation (DBP blood pressure 90 mmHg or greater, SBP 140 mmHg or greater, or both) with associated proteinuria (greater than 300 mg in 24 hours), at 24 to 33 weeks' gestation. Excluded: history of HT, chronic disorders such as diabetes, fetal malformations, or maternal or fetal comorbidities or both, use of erythromycin, ketoconazole, itraconazole, antiretroviral agents, or any other medication that could interact with sildenafil, patients using antihypertensive medications other than a‐methyldopa, not giving consent.
Interventions	Exp: oral sildenafil citrate at 50 mg every 8 hours. (50 women) Control: identical placebo capsules. (50 women)
Outcomes	Women: time interval between randomisation and delivery, MAP, Doppler velocimetry of the uterine, umbilical, and middle cerebral arteries, need for an additional antihypertensive drug, caesarean section, eclampsia, placental abruption, maternal cardiac failure, HELLP syndrome, PPH, maternal side‐effects. Babies: IUGR (< 10th centile), birthweight, Apgar score, NICU admission, days in NICU, neonatal infection, IVH, thrombocytopenia, hypotension, necrotising enterocolitis, hypoglycaemia, pneumothorax, convulsions, neonatal demise, neonatal side‐effects.
Notes	Korotkoff phase IV used for DBP. Funding: funded by university or non commercial organisation. The authors reported no potential conflicts of interest.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Blocked randomisation (blocks of 5).
Allocation concealment (selection bias)	Low risk	Sealed and numbered opaque envelopes.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Study investigators, attending physicians, ultrasonographers, neonatologists, and patients were blinded regarding allocation to placebo or sildenafil groups until the end of the study.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Study investigators, attending physicians, ultrasonographers, neonatologists, and patients were blinded regarding allocation to placebo or sildenafil groups until the end of the study.
Incomplete outcome data (attrition bias) All outcomes	Low risk	7 women (7%) discontinued treatment due to concerns about the possibility of being allocated to a placebo (2 in each group), or due to side‐effects (1 in treatment group, 2 in placebo group). However, authors stated that women consent in the use of their data and presented results in 100 women randomised.
Selective reporting (reporting bias)	Unclear risk	Assessment from published study report.
Other bias	Low risk	Groups appear comparable at baseline. Funded by university or non commercial organisation. Authors stated not potential conflict of interests.