Study	Reason for exclusion
Gmoser 1991	Inadequate reporting about method of randomisation. Inclusion and exclusion criteria were not defined. This study was carried out by Gmoser and colleague, conducted in Austria and published in 1991 (Gmoser 1991). The investigators compared length of gestation and the weight at birth, between women treated with cervical pessary versus those with no intervention. The study included 300 cases. They found that the length of gestation and birthweight were higher in the pessary group compared with no intervention (39 versus 36 weeks and 2950 versus 2400 g, respectively). They did not reach a solid conclusion from their study (Gmoser 1991).
Von Forster 1986	Using quasi‐randomisation in the form of initial of the women's surname. Unclear inclusion and exclusion criteria. Incomplete outcome data addressed, women were randomised into 3 groups and only 2 groups were included in the analysis. The study was carried out by Von Forster and colleagues in Germany. It was a prospective randomised trial conducted between 1982 and 1983. In this study, patients were randomised into 3 groups based on the initial letter of their surname (quasi‐randomised). Patients in Group 1 were admitted as in‐patients and received cerclage (n = 112). Those in Group 2 (n = 130) were fitted with a pessary as outpatients. A third group was simply ordered to rest in bed. The type of pessary used was not mentioned. The therapies were used in patients for prophylactic as well as therapeutic reasons, although these terms are not defined in the article. However, at the analysis stage the investigators only analysed the intervention groups because all patients in the rest only group did need treatment. The results reported that the 2 groups were equal in the length of pregnancy (mean 37 to 38 weeks), birthweight (mean 3000 g), Apgar scores, and fetal survival. They investigators concluded that cerclage and pessary were equally effective in the management of cervical incompetence (Von Forster 1986).