Clinical validation of risk scoring systems to predict risk of delayed bleeding after EMR of large colorectal lesions
Por:
Albeniz E, Zebenzuy Gimeno-Garcia A, Fraile M, Ibanez B, Guarner-Argente C, Alonso-Aguirre P, Antonio Alvarez M, Jerusalen Gargallo C, Pellise M, Ramos Zabala F, Herreros de Tejada A, Nogales O, Martinez-Ares D, Mugica F, de la Pena J, Espinos J, Huerta A, Alvarez A, Gonzalez-Santiago J, Navajas F, Gabriel Martinez-Cara J, Redondo-Cerezo E, Merlo Mas J, Sabado F, Rivero L, Saperas E, Soto S, Rodriguez-Sanchez J, Lopez-Roses L, Rodriguez-Tellez M, Rullan Iriarte M, Elosua Gonzalez A, Pardeiro R, Valdivielso Cortazar E, Concepcion-Martin M, Huelin Alvarez P, Colan Hernandez J, Cobian J, Santiago J, Jimenez A, Remedios D, Lopez-Viedma B, Garcia O, Martinez-Alcala F, Perez-Roldan F, Carbo J, Enguita M
Publicada:
1 abr 2020
Resumen:
Background and Aims: The Endoscopic Resection Group of the Spanish Society of Endoscopy (GSEED-RE) model and the Australian Colonic Endoscopic Resection (ACER) model were proposed to predict delayed bleeding (DB) after EMR of large superficial colorectal lesions, but neither has been validated. We validated and updated these models.
Methods: A multicenter cohort study was performed in patients with nonpedunculated lesions >= 20 mm removed by EMR. We assessed the discrimination and calibration of the GSEED-RE and ACER models. Difficulty performing EMR was subjectively categorized as low, medium, or high. We created a new model, including factors associated with DB in 3 cohort studies.
Results: DB occurred in 45 of 1034 EMRs (4.5%); it was associated with proximal location (odds ratio [OR], 2.84; 95% confidence interval [CI], 1.31-6.16), antiplatelet agents (OR, 2.51; 95% CI,.99-6.34) or anticoagulants (OR, 4.54; 95% CI, 2.14-9.63), difficulty of EMR (OR, 3.23; 95% CI, 1.41-7.40), and comorbidity (OR, 2.11; 95% CI, .99-4.47). The GSEED-RE and ACER models did not accurately predict DB. Re-estimation and recalibration yielded acceptable results (GSEED-RE area under the curve [AUC], .64 [95% CI, .54-.74]; ACER AUC, .65 [95% CI, .57.73]). We used lesion size, proximal location, comorbidity, and antiplatelet or anticoagulant therapy to generate a new model, the GSEED-RE2, which achieved higher AUC values (.69-.73; 95% CI, .59-.80) and exhibited lower susceptibility to changes among datasets.
Conclusions: The updated GSEED-RE and ACER models achieved acceptable prediction levels of DB. The GSEED-RE2 model may achieve better prediction results and could be used to guide the management of patients after validation by other external groups.
|