Sa1364 Prospective Comparison of Test Performance Between Traditional and High-Resolution Anal Manometry (HRAM) in Female Healthy Volunteers and Patients With Fecal Incontinence

Abstract

Introduction Conventional anal manometry (CAM) is widely used to measure sphincter function in patients with fecal incontinence (FI). However, results correlate poorly with symptom severity and may have limited impact on patient management. This study evaluated the discriminant utility of novel measures of anal resting and squeeze function using highresolution anorectal manometry (HRAM) compared to conventional measures. Methods Ninety-five healthy female volunteers (HV) (age 42 ± 13 years) and 95 females with FI (age 55 ± 16 years) underwent HRAM (solid state, 10 channels, 8 mm spacing). CAM data were subsampled from HRAM data. In each participant the CAM squeeze increment (CAM-SI) and CAM resting pressure (CAM-R) were compared with new derived HRAM ‘profile' measures (squeeze [HRAM-SP] and rest [HRAM-RP]). Results were compared between groups and metrics of diagnostic accuracy derived. Results Median St Marks Incontinence score was 11 (IQR=7) for FI vs. 1 (IQR=1) for HV, p <.001. In FI participants, abnormal squeeze function was demonstrated in 34/95 (36% sensitivity) using CAM-SI and 56/95 (59% sensitivity) using HRAM-SP (p <.001), resulting in superior receiver operator curve [ROC] characteristics for the novel measure (0.86 for HRAM-SP vs. 0.74 for CAM-SI). Regression demonstrated superiority of the HRAM-SP score (F = 17.91, adj R2 = 0.42) compared to CAM-SI scores (HRAM-SP: F = 12.52, adj R2 = 0.33; CAM-SI: F = 6.05, adj R2 = 0.18) in health status prediction. The difference between measures was less marked for analysis of anal resting pressure with abnormal function present in 24/95 (26% sensitivity) and 26/95 (27% sensitivity) when using CAM-R and HRAM-RP respectively. However slightly superior ROC characteristics were demonstrated for the novel measure (0.65 for HRAM-RP vs. 0.59 for CAM-R). Final cross tabulation of agreement between measures demonstrated that 25/ 61 patients who were classified as having normal function with the CAM measures displayed reduced function with wither the HRAM-R or HRAM-SP profile (26% of the cohort as a whole). Conclusions Novel HRAM measures appear superior in detecting impaired anal sphincter function in patients with FI over conventional measures and show enhanced correlation with symptom severity. In particular HRAM-SP has a 23% increase in sensitivity for detection of abnormal sphincter function in FI over conventional measures. The addition of this technique of analysis to manometric studies has a number of potential implications for clinical practice.