494: Scheduled timing of and patient compliance with longitudinal sonographic cervical length (CL) measurement in a multicenter randomized trial of ultrasound-indicated cerclage

Abstract

To assess the timing of CL measurement and compliance with scheduled visits in women at high risk for recurrent preterm birth (PTB). Women with ≥1 prior spontaneous PTB 17-336/7 wks had an initial scan at 16-21 wks. Unless randomization occurred at the initial scan, follow-up visits were scheduled biweekly if CL ≥30 mm, increased to weekly if 25-29 mm. Women were randomized: cerclage or no-cerclage if CL<25 mm. The final scan was scheduled by 226/7 wks. Birth in the prescribed scanning window 16-226/7 wks and gestational age (GA) at birth were evaluated. Of 674 non-randomized women, only 11 (1.6%) delivered in the GA window <23 wks. The rate of birth <23 wks was 1.6% in 379 women with perfect compliance and 1.7% in 295 women who were either late for any visit or who missed their final scan (p=1.00). 183 (27.2%) did not present for their final scan < 23 wks, and these non-compliant women delivered at a mean (SD) GA of 37.0 (4.0) as compared to 37.2 (3.8) wks for those who were compliant (p=0.66). Women who missed their final 1-week follow-up (last observed CL 25-29 mm), delivered at 36.4 (4.4) wks compared to 37.1 (3.9) wks in those whose missed scan was a 2-week follow-up (last observed CL ≥30 mm; p=0.55). Of 189 women randomized after the initial scan, 91 were assigned to cerclage and 98 to no-cerclage. In the cerclage group the rates of birth <23 weeks were 4% in 78 with perfect compliance and 0 (0%) in 13 who missed ≥1 scheduled visit (p=1.00). In the no-cerclage group the rates of birth <23 weeks were 7.5% in 80 with perfect compliance and 0 (0%) in 18 who were non-compliant (p=0.59). Women at high-risk for recurrent PTB followed with serial CL measurement using a defined scanning interval based on CL, with an intent for ultrasound-indicated cerclage, have a low, 1.8% risk of birth <23 wks. The effect of non compliance with scheduled scans was null. For non-randomized women who missed their final scan, mean birth GA did not depend on whether the last observed CL was 25-29 vs. ≥30 mm.