Forty-six RCTs were included in the review (n=over 40,000). Thirty-four RCTs reported adequate methods of allocation concealment. Twenty-five RCTs reported use of blinding.
In the five RCTs that reported maternal deaths there were 11 deaths: eight deaths out of 11,153 in the misoprostol groups and three deaths out of 11,125 in controls. No additional maternal deaths were reported in trial publications or identified through contact with authors.
When misoprostol was compared with other uterotonics (four RCTs), severe maternal morbidity was similar in both groups; this applied in both treatment and prevention studies. When misoprostol was compared with placebo (six RCTs), severe maternal morbidity was slightly higher in the misoprostol groups; this applied in both treatment and prevention studies. However, events were too few to draw conclusions.
Pooling of direct and adjusted indirect data showed no evidence that 600µg of misoprostol was superior to 400µg for preventing blood loss of 1,000mL or more. Significant heterogeneity (I2>50%) in the direct comparison was eliminated by exclusion of a single outlier study. Pooling of both types of data also showed that pyrexia was significantly more common in the 600µg or more misoprostol group than in the 400µg group (RR 2.53, 95% CI 1.78 to 3.60); significant heterogeneity (I2=65%) in one subgroup in the direct comparison could not be explained.
The sensitivity analysis did not materially affect the findings. Other outcomes were reported in the review.