Modified Delphi study of ultrasound signs associated with placenta accreta spectrum

Objective: To determine, by expert consensus, through a modified Delphi process the role of standardized and new ultrasound signs in the prenatal evaluation of patients at high-risk of placenta accreta spectrum (PAS). Method: A systematic review of articles providing information on the ultrasound imaging signs or markers associated with PAS was performed before the development of questionnaires for the first round of the Delphi process. Only peer-reviewed original research studies in the English language describing one or more new ultrasound signs for the prenatal evaluation of PAS were included. A three-round consensus building Delphi method was then conducted under the guidance of a steering group. The Steering group included nine experts who invited an international panel of experts in obstetric ultrasound imaging and evaluation of patients at high-risk of PAS. Strong consensus was defined as a 70% agreement between participants. Results: The systematic review identified 15 articles describing eight new ultrasound signs for the prenatal evaluation of PAS. A total of 35 external experts were approached, of whom 31 agreed and entered the first round. Thirty external experts (97%) and seven experts from the steering group completed all three rounds. A consensus was reached that a prior history of Caesarean deliveries, myomectomy or PAS should be the indication for detailed PAS ultrasound assessment. The panellists also reached a consensus that seven of the 11 conventional signs of PA, namely i) loss of the "clear zone", ii) myometrial thinning, iii) bladder wall interruption and the presence of a placental bulge, iv) exophytic mass, v) uterovesical hypervascularity, vi) placental lacunae and vii) bridging vessels should be included in the examination. A consensus was not reached for any of the eight new signs identified by the systematic review. For other ultrasound features that increase the probability of PAS at birth, the panellists reached a consensus for the finding of an anterior placenta previa or a placenta previa with cervical involvement. Only the quantification of placental lacunae using an existing score obtained a strong consensus. For predicting surgical outcome in patients with a high probability of PAS at delivery, a consensus was obtained for i) loss of the "clear zone", ii) bladder wall interruption, iii) the presence of placental lacunae and iv) a placenta previa involving the cervix. Conclusions: We have confirmed the continued importance of eight established standardised ultrasound signs of PAS, highlighted the role of TVS in evaluating the placental position and anatomy if the cervix, and identified new ultrasound signs that may become useful in the prenatal evaluation and management of patients at high-risk of PAS at birth. This article is protected by copyright. All rights reserved.