Adverse Events in Gastrointestinal Endoscopy: Validation of the AGREE Classification in a Real-life 5-year Setting

Background: The novel AGREE classification for adverse events (AEs) in gastrointestinal (GI) endoscopy has not yet been validated in a real-world setting.

Aims: Our study aims to evaluate the correlation between the grades of AEs in the ASGE and AGREE classifications and to assess the interobserver agreement of the two classification systems.

Methods: The correlation and association between the AE grades of the ASGE and AGREE classifications were analyzed using the Spearman rank correlation test and the chi-squared analysis, respectively. A weighted Cohen's kappa coefficient analysis was performed to determine the interobserver agreement of both classification systems.

Results: We prospectively collected the AEs that occurred in our endoscopy unit over the past five years. A total of 226 AEs (226/84,863, 0.3%) occurred. There was a correlation between the ASGE and AGREE classifications (ρ = 0.61) and a moderately significant association (p < 0.01, Cramer's V = 0.7). The interobserver agreement for the ASGE classification was fair (kappa 0.60, 95% confidence interval [CI]: 0.54, 0.67), whereas it was good for the AGREE classification (kappa 0.80, 95% CI: 0.62, 0.87).

Conclusions: The AGREE classification was validated for the first time in a real-world setting and showed a positive correlation and higher interobserver agreement than the ASGE classification.