METHODS
All patients with major OASIS (defined as Grade IIIb, IIIc or IV by the Sultan classification [4]) in a single birthing unit over a two year period underwent endoanal ultrasonography (B&K, Denmark) at 6-12 weeks post-surgical repair (Group A). Surgical repair was performed by a specialist consultant under anaesthesia (general or spinal) using an overlapping technique if possible, otherwise an approximation technique was used. Any ultrasonographic gap or discontinuity for any vertical length, or horizontal width, in the anal sphincters was scored as a defect. The defect was scored separately for internal anal sphincter (IAS) or external anal sphincter (EAS). The endosonographic assessment was blinded to the ultimate faecal incontinence scores obtained at a later date.
At 1-3 years after the obstetric event these patients were sent a questionnaire designed to allow extraction and calculation of the Wexner faecal incontinence score (Figure 1). We defined Wexner score of ≥ 9 as significantly affecting quality of life. [8] The Vaizey score was not used because of the integral use of ability to defer defaecation in this score, which would introduce considerable overlap with the presence of IBS complaints. However, the ability to defer defaecation longer than 15 minutes was included as an item in our questionnaire. The questionnaire also allowed us to calculate the presence and type of IBS based on the Rome IV criteria (Figure 2). Other variables that were collected were age, parity and mode of delivery, body mass index (BMI).
Two control groups were included, and these were aimed to be of equal size to Group A. The first control group consisted of primigravid women (Group B) who filled out a questionnaire at their first consultation of a midwife or obstetrician in the first trimester of their pregnancy. The second control group were women who underwent an elective Caesarean section (Group C) who filled out a questionnaire at a minimum of 12 months post-partum. Exclusion criteria were previous anal surgery, previous obstetric sphincter injury, inflammatory bowel disease and coeliac disease.
Statistical analysis was done with SPSS version 24 (IBM, USA). For comparison of binary variables Chi quare test was used, for intra-group comparison of mean (SD) Wexner scores a t-test was used and for inter-group comparison of mean (SD) Wexner scores ANOVA was used. Format of data presentation was deliberately chosen to suit subsequent clinical usefulness. The study protocol obtained ethical approval from the local health district.