OBJECTIVE: Volume measurement of the anal sphincter can be a future method for assessing volume loss, muscle atrophy or laceration. Three-dimensional (3D) endoanal ultrasound is a technique for assessing the volume of the anal sphincters, but the reproducibility of the method is scarcely known. DESIGN: Cross-sectional, repeated measurements. SAMPLE: Twenty women were recruited for the study after written consent according to an ethically approved protocol, nine 0-gravida and 11 with traumatic vaginal deliveries. METHOD: Endoanal 3D-ultrasound volume was obtained during rest and squeeze. The length of the anal canal and the volume of the external and internal sphincters were determined by two observers. Observer 1 repeated the measurements three times for all 20 women, and observer 2 for the nine 0-gravida, and intra- and inter-observer variation was assessed. RESULTS: During rest, the anal length measurement had intra-class correlation coefficients of 0.91 for observer 1 and 0.85 for observer 2. The limits of agreement for inter-observer measurement were (-0.81 to 1.61) measured in centimeters. For the external anal sphincter volume, the intra-class correlation coefficients and the limits of agreement were correspondingly: 0.89, 0.78 and (-7.29 to 6.03) measured in cm(3), and for the internal anal sphincter volume: 0.85, 0.69 and (-1.72 to 2.95). The variation in identifying the external anal sphincter could rise to a corresponding 30% error in volume measurement. CONCLUSION: Although intra-class correlation coefficients showed good reproducibility for endoanal ultrasound measurements, the limits of agreement were less reassuring with sizeable variation in the volume assessment, probably due to uncertainty in landmark identification.
OBJECTIVE: Volume measurement of the anal sphincter can be a future method for assessing volume loss, muscle atrophy or laceration. Three-dimensional (3D) endoanal ultrasound is a technique for assessing the volume of the anal sphincters, but the reproducibility of the method is scarcely known. DESIGN: Cross-sectional, repeated measurements. SAMPLE: Twenty women were recruited for the study after written consent according to an ethically approved protocol, nine 0-gravida and 11 with traumatic vaginal deliveries. METHOD: Endoanal 3D-ultrasound volume was obtained during rest and squeeze. The length of the anal canal and the volume of the external and internal sphincters were determined by two observers. Observer 1 repeated the measurements three times for all 20 women, and observer 2 for the nine 0-gravida, and intra- and inter-observer variation was assessed. RESULTS: During rest, the anal length measurement had intra-class correlation coefficients of 0.91 for observer 1 and 0.85 for observer 2. The limits of agreement for inter-observer measurement were (-0.81 to 1.61) measured in centimeters. For the external anal sphincter volume, the intra-class correlation coefficients and the limits of agreement were correspondingly: 0.89, 0.78 and (-7.29 to 6.03) measured in cm(3), and for the internal anal sphincter volume: 0.85, 0.69 and (-1.72 to 2.95). The variation in identifying the external anal sphincter could rise to a corresponding 30% error in volume measurement. CONCLUSION: Although intra-class correlation coefficients showed good reproducibility for endoanal ultrasound measurements, the limits of agreement were less reassuring with sizeable variation in the volume assessment, probably due to uncertainty in landmark identification.