ABSTRACT
A randomized study was performed in two clinics in order to compare the efficacy, safety and acceptability of a new model of a levonorgestrel-releasing intracervical device situated either in the cervical canal (group I) or in the uterine cavity (group II). Group I included 151 women (age 18-43) and group II, 147 women (age 19-43). The number of nulliparous women was 145. The total expulsion rate was 9.1%. The expulsion rate was lower among nulliparous women (4.8%) than among parous women (13.1%). There was also a significant difference in the expulsion rates between the two clinics. Two pregnancies occurred in both groups and two of these were ectopic, one in each group. The other two occurred after unnoticed expulsion. These figures give 12-month cumulative pregnancy rates of 1.3% and 1.4% in groups I and II, respectively. Total infection rate was 0.3%. The continuation rates were 82.1% in group I and 85.0% in group II after the first year. At three-month follow-up there were significantly fewer menstrual problems in group I than in group II (p = 0.0266). It is concluded that expulsion is still a problem with a small device but it may be diminished by intrauterine insertion and by selecting the users carefully.
