Roel Bogie

Impact of endoscopist training on post-colonoscopy colorectal cancer rate

accounted for approximately 2.0 per 1000 colonoscopies, 6 identical to the rate found prior to training in the present study. Since this PCCRC rate showed no significant decline over time, this observed percentage validates the pre-training cohort estimations and underlines the difference between pre-and post-training PCCRC percentages. It is however important to acknowledge that bias may have occurred. First, the follow-up ended in October 2014. Therefore, regarding the colonoscopies performed after October 2009, the occurrence of late onset PCCRCs cannot be excluded with certainty. In this way especially ‘newly developed’ PCCRCs will be missed, which is less influenced by training. To limit this bias, the rate of PCCRCs pre- and post-training were presented as PCCRCs/1000 person years of follow-up. Second, person years of follow-up prior to training were based on estimation, resulting in bias in case the real follow-up would be longer. To overcome this bias, a sensitivity analysis was performed using 60 months of follow-up in the cohort prior to training. Finally, data were prospectively collected after training, resulting in validation of the PALGA dataset since additional PCCRCs may be identified. In the retrospectively collected pre-training data this additional check was not available, theoretically leading to a small chance of underreporting PCCRCs. Taken into account these limitations, our data pointed out that focusing on endoscopist related factors such as systematic short-interval training for the recognition and complete resection of NP CRN may help to minimize PCCRC rate. In conclusion, we have shown a decrease in PCCRC rate after implementation of systematic training inNP-CRNdetection and resection bymore than 50%. The etiology of the PCCRCs liesmostly in missed lesions, and in a lesser extent to non-adherence to surveillance intervals, inadequate bowel examination and newly developed cancers. So, there is room for further improvement. In order to further minimize the PCCRC rate it should be acknowledged that a multifactorial approach is needed giving attention to training in detection and resection of polyps, especially of NP-CRNs, adherence to colonoscopy surveillance intervals and employing quality indicators for colonoscopy.

9

181