Poor agreement between operators on grading of the placenta

Area of Science:

• Obstetrics and Gynecology
• Medical Imaging
• Perinatology

Background:

• Abnormal placental grading is linked to adverse pregnancy outcomes.
• Accurate placental assessment is crucial for monitoring fetal development and health.
• Existing grading methods require evaluation for reliability.

Purpose of the Study:

• To quantify the intra- and interobserver variability in placental grading.
• To assess the reliability of the Grannum grading system among expert sonographers.
• To identify potential limitations in current placental assessment techniques.

Main Methods:

• Five expert sonographers independently graded 90 placental images on two separate occasions.
• Standardized viewing conditions were implemented, including a controlled laboratory, optimal lighting, and calibrated monitors.
• Kappa (κ) analysis was used to measure both interobserver and intraobserver agreement.

Main Results:

• Substantial variations in placental grading scores were observed among expert sonographers.
• Interobserver agreement was fair (mean κ = 0.34), with only 10% of images graded identically by all five observers.
• Intraobserver agreement was moderate (mean κ = 0.52), indicating variability even within the same observer's assessments.

Conclusions:

• The Grannum grading system for placental assessment demonstrates significant variability and lacks reliability, even among experienced professionals.
• Standardized conditions did not eliminate inconsistencies in placental grading.
• There is a critical need for developing more objective and reliable methods for assessing placental health, such as advanced software-based 2D and 3D imaging techniques.