(a) Progesterone agents versus control treatments

Quality assessment							Number of patients (%)		Absolute effect (95% CI)		Relative effect (95% CI)	Quality	Importance
No. of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Publication bias	Progesterone agents	Control	Progesterone agents	Control
Outcome: miscarriage
Progesterone versus control
7	Randomized trials	Seriousa	Not serious	Not serious	Seriousb	None	52/399 (13.0)	82/378 (21.7)	128 per 1000	217 per 1000	OR 0.53 (0.36 to 0.78)	⨁⨁◯◯ Low	Critical
									89 fewer per 1,000  (from 39 fewer to 126 fewer)
Oral dydrogesterone versus control
3	Randomized trials	Seriousa	Not serious	Not serious	Seriousb	None	30/256 (11.7)	53/235 (22.6)	112 per 1000	226 per 1000	OR 0.43 (0.26 to 0.71)	⨁⨁◯◯ Low	Critical
									114 fewer per 1,000 (from 54 fewer to 155 fewer)
Vaginal progesterone versus control
4	Randomized trials	Not serious	Not serious	Not serious	Not serious	None	22/143 (15.4)	29/143 (20.3)	155 per 1000	203 per 1000	OR 0.72 (0.39 to 1.34)	⨁⨁⨁⨁ High	Critical
									48 fewer per 1,000 (from 51 more to 113 fewer)

CI: confidence interval; OR: odds ratio; ^aeither the participants or personnel were not blinded in these studies (performance bias); ^bthe 95% CI includes appreciable harm or benefit.