Premium
A score‐based method for quality control of fetal images at routine second‐trimester ultrasound examination
Author(s) -
Salomon L. J.,
Winer N.,
Bernard J. P.,
Ville Y.
Publication year - 2008
Publication title -
prenatal diagnosis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.956
H-Index - 97
eISSN - 1097-0223
pISSN - 0197-3851
DOI - 10.1002/pd.2016
Subject(s) - sonographer , medicine , reproducibility , ultrasound , nuclear medicine , radiology , cohen's kappa , quality score , mathematics , statistics , metric (unit) , operations management , economics
Objectives Our aim was to develop and evaluate the feasibility and reproducibility of score‐based quality control for routine standardized fetal ultrasound images obtained in the second trimester of pregnancy. Study Design In France, a minimum of three biometrical and six anatomical standardized ultrasound planes are to be produced with any mid‐trimester scan. All anatomical standardized ultrasound images, routinely obtained by one trained operator at 20 to 24 weeks, were stored prospectively during a 1‐year period. Twenty‐five examinations containing these images were later randomly selected. These were then analyzed by two reviewers, according to predefined criteria agreed upon on the basis of established standards. This yielded a total score of up to 32 points. Feasibility, inter‐ and intra‐reviewer reproducibility were analyzed. Results Routine second‐trimester ultrasound examinations numbering 1160 performed over a one year period by one trained sonographer unaware of the subsequent study at the time the images were recorded and stored in a database. Among the 150 images randomly selected and analyzed, adjusted kappa values were above 0.8 for 27 (84%) and 30 (94%) criteria, intra‐class correlation coefficient was 0.86 (0.75; 0.96) and 0.98 (0.94; 1) and the mean difference (95% CI) in score was − 0.44 (−3.0; 2.1) and − 0.2(−2; 1.6) for inter‐ and intra‐reviewer comparisons respectively. Conclusion A quality control policy based on image scoring is feasible and allows for good inter‐ and intra‐reviewer reproducibility. Besides its potential for audit and quality control, this could also be useful during the training process. Copyright © 2008 John Wiley & Sons, Ltd.