Bowel Preparation Scales: When Glasses Become More Important Than the View: Presidential Poster

Introduction: Clinical trials assessing efficacy of colon preparations have relied on a number of validated scales varying in complexity from a simple global rating scale to more complex composite scale with segmental assessment. Few trials have examined inter-and intraobserver variability to confirm reproducibility and concordance between examiners or between scales. These trials have suffered important limitations including use of few standardized videos, truncated recordings, and assessment in a research setting where only a specific scale range is selected. We aimed to evaluate the validity, reliability, and concordance of the 3 most commonly used bowel preparation scales in assessing colon cleansing. Methods: Elective colonoscopies were prospectively examined in the context of a clinical trial. Quality was scored by 4 certified gastroenterologists and 2 senior fellows using modified Aronchick scale, Ottawa bowel preparation scale, and Boston bowel preparation scale. Agreement on adequacy between scales was performed using the modified kappa (k) statistic. Adequacy was defined as excellent or good for the Aronchick, ≤7 for the Ottawa, and ≥5 for Boston. To assess intra and interrater variability, 20 procedures were recorded. Nontruncated videos were subsequently watched ≥1 month later and graded by the 6 endoscopists. To standardize comparison across all scores to 5 categories each, Boston and Ottawa scores were uniformly re-categorized 2:1 and 3:1, respectively. Weighed k testing was used to measure agreement. Krippendorff ’s alpha intraclass correlation coefficient (α) measured concordance with values <0.4 interpreted as fair to poor, 0.4-0.6 as moderate, 0.6-0.8 as substantial, and >0.8 as excellent agreement. Results: One hundred ninety-eight colonoscopies were assessed using the Aronchick, Ottawa, and Boston scales. Agreement between the 3 scales was poor overall (α=0.32; 95% CI 0.03, 0.57) with pairwise concordance of: Ottawa and Boston (poor, α =0.08), Aronchick and Boston (moderate, α =0.51), and Aronchick and Ottawa (fair, α=0.37). Interrater agreement was substantial for Boston (α = 0.63; 95% CI 0.55, 0.70) and moderate for Aronchick (α=0.59; 95% CI 0.52, 0.66) and Boston (α =0.58; 95% CI 0.51, 0.65). Intrarater agreement was substantial in case of Ottawa (k=0.61±0.14; p<.001), and poor for both Aronchick (k=0.36±0.15; p=0.005), and Boston (k=0.34±0.15; p=0.03). Of note, when discordance occurred it did not exceed 1 category. Conclusion: Currently used preparation adequacy scales vary widely and exhibit substantial discordance. This is likely due to quality of descriptions and the multiple utilized item decision-making processes. A new bowel preparation scale that reliably and reproducibly defines adequacy of preparations is needed.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.