We were very interested to read the randomised study by Ngai et al. comparing oral misoprostol with intravenous oxytocin for the induction of labour in women with prelabour rupture of the membranes at term (Vol 107, February 2000)1. Although the trial was well conducted in several aspects, including a single investigator performing all the vaginal examinations and the use of intrauterine pressure catheters early in labour, the analysis of the data is questionable and some conclusions must be challenged.

Stratification according to parity is necessary if both nulliparous and multiparous women are to be studied together, but analysis according to parity is of little value with such small numbers of multi-parae. What are the results if all women are analysed together (which is presumably what the power calculation is based on)?

Although several larger studies using vaginal misoprostol have shown significant increases in the incidence of tachysystole2, this small study1 as well as the previous study by the same authors3 failed to demonstrate a significant difference using an oral preparation (32.5%vs. 25%). The only other published study using oral misoprostol2 does not comment on incidence of tachysystole, apart from saying that no intervention was necessary for uterine overstimulation.

Although 40% of the women required more than one dose of misoprostol, no mention is made of the number of doses in the women with tachysystole. The demonstration of the maximum effect of misoprostol at five hours led to the suggestion that increasing the dosage interval of misoprostol may decrease the incidence of tachysystole. In our current trial where 145 women have been induced with 50 μg intravaginal misoprostol, tachysystole or hyperstimulation occurred in 41 women (28%) of whom 30 (73%) had received only a single dose of misoprostol. The suggestion of increasing the dose interval may therefore be of limited help. The absence of any cases of hyper-stimulation in the study by Ngai et al. could well be due to the small numbers.

This trial had an 80% power to detect a different of 240 minutes in the induction-delivery time. To make any comment on caesarean section rates in the two groups with a total of five caesarean sections is impossible. A 90% power of detecting a decrease in caesarean section rate by 25% (from 10% to 7.5%)2 would require 2500 women in each arm. No trial published so far has been large enough to comment reliably on the effect of induction of labour with misoprostol on the caesarean section rate or other less common maternal or neonatal outcomes, and we would agree that further evaluation is required.


  1. Top of page
  2. References