ObjectiveThe objectives were to assess whether anatomical location of ultrasound (USS) indicated cervical cerclage and/or the degree of cervical shortening (cervical length; CL) prior to and following cerclage affects the risk of preterm birth (PTB).MethodA retrospective cohort study of 179 women receiving cerclage for short cervix (≤25mm) was performed. Demographic data, CL before and after cerclage insertion, height of cerclage (distance from external os) and gestation at delivery were collected. Relative risk (RR) and odds ratio (OR) of preterm delivery were calculated according to the anatomical location of the cerclage within the cervix and the CL before and after cerclage as categorical and continuous variables. Partition tree analysis was used to identify the threshold cerclage height that best predicts PTB.Results25% (n = 45) delivered <34 weeks and 36% (n = 65) delivered <37 weeks. Risk of PTB was greater with cerclage in the distal 10mm (RR2.37, 95% CI 1.45–3.87) or the distal half of a closed cervix (RR2.16, 95% CI 1.45–3.87). Increasing absolute cerclage height was associated with a reduction in PTB (OR 0.87, 95% CI 0.82–0.94). A cerclage height <14.5 mm best predicts PTB (70.8%). Increasing CL following cerclage was associated with a reduction in PTB (OR0.87, 95% CI 0.82–0.94). Conversely, the risk of PTB was increased where CL remained static or shortened further following cerclage (RR2.34, 95% CI 1.04–5.25).ConclusionThe higher a cerclage was placed within a shortened cervix, the lower the subsequent odds of PTB. Women whose cerclage is placed in the distal 10mm of closed cervix or whose cervix fails to elongate subsequently, should remain under close surveillance as they have the highest risk of PTB.