Do we need individualized interslice intervals for exoanal tomographic imaging?

OBJECTIVE

To compare a standardized interslice interval of 2.5 mm against individualized interslice intervals for exoanal sphincter imaging, in the prediction of anal incontinence, visual analogue scale (VAS) bother score of anal incontinence and St Mark's incontinence score.

METHODS

This was a cross-sectional study of 1342 women seen between December 2017 and May 2021 for symptoms of pelvic floor dysfunction. All women underwent an interview, including St Mark's score and VAS bother score for anal incontinence, as well as tomographic exoanal sphincter imaging. Tomographic coronal plane slices were obtained by a sonographer blinded to all clinical data, first using individualized interslice intervals of 1.5-4.5 mm and then a second time using an interval of 2.5 mm, performed successively to ensure blinding. Multivariable ordinal and linear regression were employed to assess the association of 'significant residual defects' of the EAS with symptoms of anal incontinence, St Mark's score and VAS bother score of anal incontinence. Cohen's κ was used to determine agreement between the two methods.

RESULTS

Of 1342 women seen during the inclusion period, ultrasound volume data were missing in 211 and information on anal incontinence was missing in three, leaving 1128 complete datasets for analysis. Anal incontinence was reported by 253 (22.4%) women, with a mean St Mark's score of 11 (range, 1-22) and a mean VAS bother score of 6 (range, 0-10). Individualization of interslice intervals resulted in a mean ± SD interval of 2.7 ± 0.5 mm (range, 1.5-4.5 mm). There were 429 (38.0%) women who had at least one abnormal external anal sphincter (EAS) slice, and a 'significant residual defect' of the EAS was diagnosed in 138 (12.2%). Reanalysis using a standardized interval of 2.5 mm resulted in abnormal EAS slices in 461 (40.9%) women, with 'significant residual defects' in 134 (11.9%) women. Of the 1128 women, only 28 (2.5%) were classified differently when a 2.5-mm interval was used; agreement between the two methods was high (κ = 0.88). The unadjusted and adjusted models for individualized and standardized interslice intervals did not differ significantly for association with anal incontience.

CONCLUSION

A standardized interslice interval of 2.5 mm does not seem to be less valid for the diagnosis of 'significant residual EAS defect' than individualized interslice intervals. This will help to simplify automated analysis of sphincter imaging. © 2025 International Society of Ultrasound in Obstetrics and Gynecology.