A multicentre validation of the BBPS bowel preparation grading score by UK endoscopists; time to upgrade to segmental scoring

Background

Adequate bowel preparation is required to perform good quality colonoscopy. In the UK a categorical grading of the entire colon preparation quality is recommended as part of key performance indicators. The Boston Bowel Preparation Scale (BBPS) is a grading scale that divides the colon into three segments, with each having an individual score. A higher BBPS correlates more closely with greater polyp detection than categorical grading scales. This study assesses the interobserver reliability of the BBPS in a cohort of UK endoscopists.

Methods

This was a prospective observational study assessing the interobserver reliability of the BBPS scale. All endoscopists who took part in the EBOPS (an RCT of the effect of an educational video on bowel preparation) study were invited to participate. Videos of the withdrawal of 10 colonoscopy examinations were created. Participants completed BBPS and the adequacy of the preparation to detect >5mm polyps, flat polyps, or depressed polyps. The ICC for interobserver reliability was compared for total and segmental BBPS scores.

Results

27 endoscopists participated in the study. 20/27 (74%) were consultants, 22% were clinical endoscopists and 4% were fellows. The median annual colonoscopies completed were 150 (IQR = 190) with a range of 50–600. The inter-rater ICC for segmental BBPS score was near perfect at 0.987 (95% CI 0.978 - 0.993). The inter-rater ICC for total BBPS score was near perfect at at 0.988 (95% CI 0.974 - 0.997). In colonoscopy videos assessed to be adequate (BBPS ≥ 2 in each segment) the preparation was deemed sufficient to detect a polyp >5mm in size in 89%, a flat polyp in 85%, and a depressed polyp in 79%. Whereas, when the bowel preparation was inadequate, the observers felt the preparation was adequate to detect >5mm polyps in 24%, flat polyps in 5%, and depressed polyps in 3%.

Conclusions

This study demonstrates the interobserver reliability of the BBPS grading scale for the first time in UK-based endoscopists. With near perfect correlation demonstrated for both segmental and total BBPS scores, it indicates the consistency of this grading scale within this population. The subjective likelihood of adequacy of bowel preparation for polyp detection (>5mm, flat and depressed) increased with both total BBPS score and adequate preparation. The perceived likelihood of polyp detection is in keeping with published evidence of correlation of BBPS score with ADR. Since the interobserver reliability has been demonstrated in a group of endoscopists of differing experience, and the BBPS scale correlates more closely with polyp detection than categorical grading, is it not time that it’s use became standard practice within the UK?