WILL RANDOM 100 PROCEDURES BE ENOUGH TO ESTIMATE THE COLONOSCOPY QUALITY METRICS OF A FULL COLORECTAL CANCER SCREENING SETTING?

Aims Recent guidelines suggest the possibility of evaluation of 100 procedures to estimate colonoscopy quality metrics in a cohort instead of a complete evaluation of the all cohort. We intend to compare both approaches in a colorectal cancer screening cohort.

Methods Prospective cohort study for evaluation of quality metrics in colonoscopies performed in a single unit in 2018-2019 with screening indication. Evaluation of quality metrics according ESGE guideline in the full cohort and in 100 procedures’ sample. Recorded, demographics, Bowel preparation, polyp (PDR) and adenoma (ADR) detection rate, Cecal intubation rate (CIR), withdrawal time, polypectomy technic, complications and post-colonoscopy orientation. Automatic generated sample using SPSS. Statistical analysis with Qui² and t-test.

Results 800 colonoscopies in the total cohort and a random sample of 100 procedures were compared. Comparing the 800 cohort vs. 100 sample: no differences regarding sex (male: 52.5% vs. 52%, p=0.3); mean age (62.5 vs 62.6, p=0.81) and risk factors for inadequate bowel preparation. Adequate bowel preparation was 91.5% vs. 90% (p= 0.61), with mean BBPS 7.37 vs. 7.14 (p=0.17); a CIR of 95.7% vs. 95% (p=0.75) and a mean withdrawal time of 8.3 vs. 7.9 minutes (p=0.79). The PDR 52.5% vs 52% (p= 0.93) and ADR 40.7% vs. 37% (p=0.48) were also similar. Appropriate polypectomy technique 97% vs. 96% (p=0.59); Appropriate lesion description 93 vs. 89% (p=0.15); Polyp retrieval rate 93.6% vs. 94.1% (p=0.85); Complications 0.8% vs. 2% (p=0.24), appropriate post-polypectomy surveillance recommendations 97.4% vs. 98% (p=0.72) and patients experience reported as very good 80.6% vs. 81.6% (p=0.84) were again similar.

Conclusions In our colorectal screening cohort, the use of just a 100-sample of procedures offers enough accuracy to infer on the quality metrics of the full colonoscopy cohort and might be used as a resource solution in cases where the full cohort cannot be assessed.