Fourteen RCTs (641 women) were included.
Ephedrine was given before hypotension developed in 13 of the 14 RCTs.
Allocation concealment was adequate (grade A) in 4 RCTs and unclear (grade B) in 9 RCTs. It could not be clarified whether one trial was randomised. Seven RCTs were double-blinded, 4 were single-blinded and 3 were not blinded. Three RCTs provided details about withdrawals. Sample size calculations had been performed in 3 RCTs.
There was a wide variation in the ephedrine regimens used in the studies.
Maternal outcomes.
Hypotension (12 RCTs, 571 women): significantly fewer women experienced hypotension with ephedrine, compared with control (RR 0.73, 95%: CI 0.63, 0.86). There was no evidence of statistical heterogeneity (P=0.24). The number-needed-to-treat, assuming a baseline risk of 80%, was 4.6 (95% CI: 3.4, 8.9).
The overall effect was similar after excluding one RCT with uncertainty regarding the method of randomisation (RR 0.74, 95% CI: 0.65, 0.85), and was unchanged (RR 0.73, 95% CI: 0.61, 0.86) after excluding one RCT in which ephedrine was not given unless the arterial blood-pressure decreased to below baseline. The results were similar for the 4 RCTs with adequate allocation concealment (RR 0.69, 95% CI: 0.58, 0.83) and for the 7 RCTs with unclear allocation concealment (RR 0.76, 95% CI: 0.59, 0.99); in both cases there was no evidence of heterogeneity (p=0.56 and p=0.12, respectively). Double-blind RCTs found that significantly fewer women experienced hypotension with ephedrine, compared with control, but for single-blind RCTs there was no statistically significant difference between the two groups: the RRs were 0.70 (95% CI: 0.57, 0.87) and 0.79 (95% CI: 0.58, 1.08) for the 7 double-blind RCTs and 4 single-blind RCTs, respectively.
Maternal hypertension (5 RCTs with specific definitions of reactive hypertension): there was no significant difference between ephedrine and control in the number of women with reactive hypertension (RR 1.63, 95% CI: 0.93, 2.84).
Abnormal maternal heart rate: there was no significant difference between ephedrine and control in the number of women with tachycardia (greater than 120 beats/minute) in one RCT with 40 women (14 out of 20 with ephedrine versus 13 out of 20 with placebo). There was no significant difference between ephedrine and control in the number of women with bradycardia in the 2 RCTs that reported maternal bradycardia (RR 0.52, 95% CI: 0.04, 5.96).
Nausea and vomiting: there was no significant difference between ephedrine and control in the number of women with nausea, vomiting or combined nausea and vomiting. The RR was 0.82 (95% CI: 0.57, 1.18) for nausea (3 RCTs), 0.73 (95% CI: 0.35, 1.52) for vomiting (3 RCTs), and 0.71 (95% CI: 0.37, 1.37) for nausea and vomiting (6 RCTs).
Neonatal outcomes.
Low Apgar scores at 1 minute: there was no significant difference between ephedrine and control in the number of neonates with a low Apgar score (RR 0.77, 95% CI: 0.29, 2.06).
Umbilical pH (8 RCTs, 301 women): statistically significant heterogeneity was found (p=0.03). The mean umbilical pH ranged from 7.23 to 7.29.
Foetal acidosis (6 RCTs, n=350): there was no significant difference between ephedrine and control in the number of neonates with acidosis (RR 1.36, 95% CI: 0.55, 3.35).
Standard base excess: there was no significant difference between ephedrine and control for arterial base excess. The weighted mean difference (3 RCTs, n=110) was -0.85 (95% CI: -2.32, +0.61).
There was no evidence of publication bias in the funnel plot (intercept p=0.41).