• Ziad A Haidar, Ramesha Papanna, Baha M Sibai, Nina Tatevian, Oscar A Viteri, Patricia C Vowels, Sean C Blackwell, and Kenneth J Moise.
• Department of Obstetrics, Gynecology, and Reproductive Sciences, McGovern Medical School at The University of Texas Health Science Center at Houston, Houston, TX. Electronic address: ziad.haidar1@gmail.com.
• Am. J. Obstet. Gynecol. 2017 Aug 1; 217 (2): 202.e1-202.e13.

BackgroundTraditionally, 2-dimensional ultrasound parameters have been used for the diagnosis of a suspected morbidly adherent placenta previa. More objective techniques have not been well studied yet.ObjectiveThe objective of the study was to determine the ability of prenatal 3-dimensional power Doppler analysis of flow and vascular indices to predict the morbidly adherent placenta objectively.Study DesignA prospective cohort study was performed in women between 28 and 32 gestational weeks with known placenta previa. Patients underwent a two-dimensional gray-scale ultrasound that determined management decisions. 3-Dimensional power Doppler volumes were obtained during the same examination and vascular, flow, and vascular flow indices were calculated after manual tracing of the viewed placenta in the sweep; data were blinded to obstetricians. Morbidly adherent placenta was confirmed by histology. Severe morbidly adherent placenta was defined as increta/percreta on histology, blood loss >2000 mL, and >2 units of PRBC transfused. Sensitivities, specificities, predictive values, and likelihood ratios were calculated. Student t and χ2 tests, logistic regression, receiver-operating characteristic curves, and intra- and interrater agreements using Kappa statistics were performed.ResultsThe following results were found: (1) 50 women were studied: 23 had morbidly adherent placenta, of which 12 (52.2%) were severe morbidly adherent placenta; (2) 2-dimensional parameters diagnosed morbidly adherent placenta with a sensitivity of 82.6% (95% confidence interval, 60.4-94.2), a specificity of 88.9% (95% confidence interval, 69.7-97.1), a positive predictive value of 86.3% (95% confidence interval, 64.0-96.4), a negative predictive value of 85.7% (95% confidence interval, 66.4-95.3), a positive likelihood ratio of 7.4 (95% confidence interval, 2.5-21.9), and a negative likelihood ratio of 0.2 (95% confidence interval, 0.08-0.48); (3) mean values of the vascular index (32.8 ± 7.4) and the vascular flow index (14.2 ± 3.8) were higher in morbidly adherent placenta (P < .001); (4) area under the receiver-operating characteristic curve for the vascular and vascular flow indices were 0.99 and 0.97, respectively; (5) the vascular index ≥21 predicted morbidly adherent placenta with a sensitivity and a specificity of 95% (95% confidence interval, 88.2-96.9) and 91%, respectively (95% confidence interval, 87.5-92.4), 92% positive predictive value (95% confidence interval, 85.5-94.3), 90% negative predictive value (95% confidence interval, 79.9-95.3), positive likelihood ratio of 10.55 (95% confidence interval, 7.06-12.75), and negative likelihood ratio of 0.05 (95% confidence interval, 0.03-0.13); and (6) for the severe morbidly adherent placenta, 2-dimensional ultrasound had a sensitivity of 33.3% (95% confidence interval, 11.3-64.6), a specificity of 81.8% (95% confidence interval, 47.8-96.8), a positive predictive value of 66.7% (95% confidence interval, 24.1-94.1), a negative predictive value of 52.9% (95% confidence interval, 28.5-76.1), a positive likelihood ratio of 1.83 (95% confidence interval, 0.41-8.11), and a negative likelihood ratio of 0.81 (95% confidence interval, 0.52-1.26). A vascular index ≥31 predicted the diagnosis of a severe morbidly adherent placenta with a 100% sensitivity (95% confidence interval, 72-100), a 90% specificity (95% confidence interval, 81.7-93.8), an 88% positive predictive value (95% confidence interval, 55.0-91.3), a 100% negative predictive value (95% confidence interval, 90.9-100), a positive likelihood ratio of 10.0 (95% confidence interval, 3.93-16.13), and a negative likelihood ratio of 0 (95% confidence interval, 0-0.34). Intrarater and interrater agreements were 94% (P < .001) and 93% (P < .001), respectively.ConclusionThe vascular index accurately predicts the morbidly adherent placenta in patients with placenta previa. In addition, 3-dimensional power Doppler vascular and vascular flow indices were more predictive of severe cases of morbidly adherent placenta compared with 2-dimensional ultrasound. This objective technique may limit the variations in diagnosing morbidly adherent placenta because of the subjectivity of 2-dimensional ultrasound interpretations.Copyright © 2017 Elsevier Inc. All rights reserved.

29 keywords...

hide…