Study	Reason for exclusion
Abbassi 2008	Not a randomised trial ‐ quasi‐experimental study.
Ascher‐Walsh 2000	This study compared outpatient cervical ripening regimens at 40‐41 weeks' gestation. 100 mcg oral misoprostol, 200 mcg oral misoprostol or placebo were given every 3 days until 42 weeks. At 42 weeks labour was induced either with oxytocin or vaginal dinoprostone. This protocol differs substantially from the standard protocols, i.e. its primary aim is to achieve spontaneous onset of labour. The aim of the protocols included in this review is to achieve vaginal birth quickly and safely.
Ayaz 2008	Not a randomised trial, quasi‐experimental study.
Bozhinova 2007	Women who received vaginal misoprostol were also given sub‐lingual misoprostol at a dose of 50 mg every 4 hours "til the regular birth activity was reached".
Bricker 2008	In the intervention group oral misoprostol was preceded by a dose of vaginal misoprostol.
Delaney 2001	Abstract only. Women in both groups were subjected to amniotomy and the intervention was only introduced for those not in labour 1 hour later.
Hassan 2005	Not a randomised trial ‐ alternate women allocated to each group.
Ho 2010	The objective of the study was to compare titrated oral misoprostol to intravenous oxytocin for labour augmentation among women at 36 to 42 weeks of gestation with spontaneous onset of active labour.
Kadanali 1996	In this study, the initial dose of misoprostol (100 mcg) was administered vaginally followed by oral administration (100 mcg every 2 hours). This study is included in the Cochrane review on vaginal misoprostol.
Neto 1988	In this study, 15 women were divided in three groups: (i) oral misoprostol (400 mcg every 4 hours), (ii) oral misoprostol (200 mcg every 4 hours) and (iii) vaginal misoprostol (200 mcg once). The authors reported only outcomes related to the uterine activity, i.e. administration to contractions interval and strength and duration of uterine contractions.
Rasheed 2007 (V50)	This study of 310 women included 25 non‐randomised women in one arm. These women were using the unit's standard protocol (which was the same as the oral misoprostol arm study protocol) at the start of the study and so their data were included in the results. There is no analysis available without the inclusion of these non‐randomised participants.
Robinson 2011	Ongoing study designed as non‐randomised study.
Thigpen 2004	Abstract only. Vaginal misoprostol compared with oral misoprostol combined with transcervical Foley catheter. Study to be included in mechanical methods review.
Windrim 1997	The comparator group in this study was managed according to the hospital's established induction protocol. This meant that women in the comparator group were induced either with intracervical dinoprostone (0.5 mg) or intravaginal dinoprostone 1 mg every 6 hours, or intravaginal dinoprostone 2 mg every 6 hours, or dilute oxytocin infusion. The exact numbers of women per method were not reported. In addition, 11 women in the comparator group were induced with vaginal misoprostol (50 mcg every 5 hours). We felt that the interventions in the comparator group were sufficiently different to be 'lumped' together.
Zvandasara 2008	In addition to live fetuses the study also included 4 cases of IUFD (1 in the oral misoprostol group and 3 in the vaginal misoprostol group).

IUFD: intrauterine fetal death
 mcg: micrograms
 mg: milligram
 PROM: prelabour rupture of membranes