Prenatal ultrasound staging system for placenta accreta spectrum disorders

Objectives: To develop a prenatal ultrasound staging system for placenta accreta spectrum (PAS) disorders in women with placenta previa and to evaluate its association with surgical outcome, placental invasion and the clinical staging system for PAS disorders proposed by the International Federation of Gynecology and Obstetrics (FIGO). Methods: This was a secondary retrospective analysis of prospectively collected data from women with placenta previa. We classified women according to the following staging system for PAS disorders, based upon the presence of ultrasound signs of PAS in women with placenta previa: PAS0, placenta previa with no ultrasound signs of invasion or with placental lacunae but no evidence of abnormal uterus–bladder interface; PAS1, presence of at least two of placental lacunae, loss of the clear zone or bladder wall interruption; PAS2, PAS1 plus uterovescical hypervascularity; PAS3, PAS1 or PAS2 plus evidence of increased vascularity in the inferior part of the lower uterine segment potentially extending into the parametrial region. We explored whether this ultrasound staging system correlates with surgical outcome (estimated blood loss (EBL, mL), units of packed red blood cells (PRBC), fresh frozen plasma (FFP) and platelets (PLT) transfused, operation time (min), surgical complications defined as the occurrence of any damage to the bladder, ureters or bowel, length of hospital stay (days) and admission to intensive care unit (ICU)) and depth of placental invasion. The correlation between the present ultrasound staging system and the clinical grading system proposed by FIGO was assessed. Prenatal and surgical management were not based on the proposed prenatal ultrasound staging system. Linear and multiple regression models were used. Results: Two-hundred and fifty-nine women were included in the analysis. Mean EBL was 516 ± 151 mL in women with PAS0, 609 ± 146 mL in those with PAS1, 950 ± 190 mL in those with PAS2 and 1323 ± 533 mL in those with PAS3, and increased significantly with increasing severity of PAS ultrasound stage. Mean units of PRBC transfused were 0.05 ± 0.21 in PAS0, 0.10 ± 0.45 in PAS1, 1.19 ± 1.11 in PAS2 and 4.48 ± 2.06 in PAS3, and increased significantly with PAS stage. Similarly, there was a progressive increase in the mean units of FFP transfused from PAS1 to PAS3 (0.0 ± 0.0 in PAS1, 0.25 ± 1.0 in PAS2 and 3.63 ± 2.67 in PAS3). Women presenting with PAS3 on ultrasound had significantly more units of PLT transfused (2.37 ± 2.40) compared with those with PAS0 (0.03 ± 0.18), PAS1 (0.0 ± 0.0) or PAS2 (0.0 ± 0.0). Mean operation time was longer in women with PAS3 (184 ± 32 min) compared with those with PAS1 (153 ± 38 min) or PAS2 (161 ± 28 min). Similarly, women with PAS3 had longer hospital stay (7.4 ± 2.1 days) compared with those with PAS0 (3.4 ± 0.6 days), PAS1 (6.4 ± 1.3 days) or PAS2 (5.9 ± 0.8 days). On linear regression analysis, after adjusting for all potential confounders, higher PAS stage was associated independently with a significant increase in EBL (314 (95% CI, 230–399) mL per one-stage increase; P < 0.001), units of PRBC transfused (1.74 (95% CI, 1.33–2.15) per one-stage increase; P < 0.001), units of FFP transfused (1.19 (95% CI, 0.61–1.77) per one-stage increase; P < 0.001), units of PLT transfused (1.03 (95% CI, 0.59–1.47) per one-stage increase; P < 0.001), operation time (38.8 (95% CI, 31.6–46.1) min per one-stage increase; P < 0.001) and length of hospital stay (0.83 (95% CI, 0.46–1.27) days per one-stage increase; P < 0.001). On logistic regression analysis, increased severity of PAS was associated independently with surgical complications (odds ratio, 3.14 (95% CI, 1.36–7.25); P = 0.007), while only PAS3 was associated with admi