A randomized controlled trial of outpatient versus inpatient labour induction with vaginal controlled-release prostaglandin-E2: effectiveness and satisfaction.

Background: Outpatient management in obstetrics is expanding, but evidence to support outpatient labour induction is needed.

Objective: To compare the effectiveness, acceptability, duration of hospitalization, and safety of outpatient and inpatient induction of labour with intravaginal controlled-release prosta-glandin-E2 (CR-PGE2).

Methods: A prospective, randomized, controlled trial enrolled 300 women at term with parity < or = 5 and singleton pregnancies in cephalic presentation. Each had an unscarred uterus, a normal non-stress test (NST), and a Bishop score of < or = 6. After insertion of the CR-PGE2, and 1 hour of monitoring, those in the outpatient group were discharged home, to return with onset of labour or 12 hours later for an NST. If not already in labour 24 hours later, the women returned for inpatient induction. Vaginal examination was not repeated before 24 hours unless the patient was contracting and required analgesia. Inpatients remained on the antepartum ward but were otherwise treated similarly. The women in both groups reported ratings of satisfaction, pain, and anxiety over the telephone until they were in labour.

Results: There were 150 women randomized to outpatient and 150 women to inpatient induction of labour. The number of women who were in labour or who delivered by 24 hours in the outpatient group was 115 (0.77, 95% confidence interval [CI] 0.70-0.84) and in the inpatient group was 107 (0.72, 95% CI 0.64-0.79). The median times to labour were 9.8 hours (95% CI, 8.1-11.4) and 11.4 hours (95% CI, 10.1-12.7), and to delivery were 21.4 hours (95% CI, 19.2-23.5) and 20.7 hours (95% CI, 18.4-23.0), for the outpatient and inpatient groups, respectively. In the outpatient group, 56% of women reported high satisfaction during the initial 12 hours of induction compared to 39% in the inpatient group (p < 0.008). Ratings of pain and anxiety during the first 12 hours of induction were similar. In the outpatient group, women were at home for a median of 8 hours (95% CI, 6.7-9.4) before labour and delivery. There were no significant differences in adverse outcomes.

Conclusions: This study suggests that outpatient induction of labour with intravaginal CR-PGE2 may be a reasonable option for selected low-risk women; however, further study is needed to confirm the safety of this approach.