Interobserver variability in MRI assessment of the severity of placenta accreta spectrum disorders: Interobserver variability in MRI assessment of PAS

OBJECTIVE: To evaluate the level of agreement in the prenatal magnetic resonance imaging (MRI) assessment of the severity of placenta accreta spectrum (PAS) disorders in centers with high expertise in their diagnosis and management. METHODS: The MRI scans of women at risk of PAS were retrieved from the hospital electronic database and assessed by four different experts who were blinded to the final diagnosis. Each examiner was asked to judge the MRI scans according to the presence, depth and topography of placental invasion. The depth of invasion was defined as the degree of adhesion and invasion of the placenta through the myometrium and uterine serosa at the histopathological examination of the removed uterus (accreta, increta and percreta), while topography as the site of placental invasion within the uterus. The degree of inter-rater agreement (IRA) in calculating both the percentage of observed agreement among raters and the Fleiss kappa were used to analyze the data. RESULTS: Forty-six women were included in the study. There was an excellent agreement among the four different examiners in the assessment of the overall presence of PAS disorders (IRA: 92.1, 95% CI 86.8-94.0; K: 0.90, 95% CI 0.89-1). However, there was significant heterogeneity in the agreement among the different examiners when assessing the different MRI signs suggestive of PAS. There was also an excellent agreement in the identification of the depth (IRA: 98.9, 95% CI 96.8-100; K: 0.95, 95% CI 0.89-1.0) of PAS disorders. Conversely, the agreement in assessing the topography of placental invasion was only moderate (IRA: 72.8, 95% CI 72.7-72.9; K: 0.56, 95% CI 0.54-0.66). More importantly, when assessing parametrial invasion, which is one of the most significant prognostic factors in women affected by PAS, the agreement was moderate in judging the presence of invasion on coronal (IRA: 86.6%. 95% CI 86.5-86.7, K-: 0.69, 95% CI 0.59-0.71) and axial (IRA: 78.6, 95% CI 78.5-78.7, K: 0.56, 95% CI 0.33-0.60) planes. Likewise, the agreement in judging the presence and the number of newly formed vessels in the parametrial tissue was moderate (IRA: 88.0, 95% CI 88.0-88.1, K: 0.59, 95% CI 0.45-0.68) and weak (IRA: 66.7, 95% CI 66.6-66.7, K: 0.22, 95% CI 0.12-0.37), respectively. CONCLUSION: MRI has excellent inter-observer variability in detecting the presence and depth of placental invasion, while the agreement in describing the topography of invasion is less. The findings from this study highlight the need for a standardized MRI staging system of PAS disorders, in order to facilitate objective correlation between prenatal imaging, pregnancy outcomes and surgical management. This article is protected by copyright. All rights reserved.