Endoscopy 2018; 50(04): S27
DOI: 10.1055/s-0038-1637106
ESGE Days 2018 oral presentations
20.04.2018 – Colon: Improving characterization
Georg Thieme Verlag KG Stuttgart · New York

A HIGH PROFICIENCY ON OPTICAL DIAGNOSIS WITH NARROW-BAND IMAGING (NBI) CAN BE ACHIEVED BY A NON-GUIDED SELF-LEARNING PROGRAM

M Bustamante Balén
1   Hospital Universitari i Politècnic La Fe, Endoscopy Research Group. Gastroenterology Department, Valencia, Spain
,
C Satorres Paniagua
1   Hospital Universitari i Politècnic La Fe, Endoscopy Research Group. Gastroenterology Department, Valencia, Spain
,
N Alonso Lazaro
1   Hospital Universitari i Politècnic La Fe, Endoscopy Research Group. Gastroenterology Department, Valencia, Spain
,
MT Blazquez Martinez
2   Hospital Universitari i Politècnic La Fe, Gastroenterology Department, Valencia, Spain
,
A Nevarez Heredia
1   Hospital Universitari i Politècnic La Fe, Endoscopy Research Group. Gastroenterology Department, Valencia, Spain
,
M Ponce Romero
1   Hospital Universitari i Politècnic La Fe, Endoscopy Research Group. Gastroenterology Department, Valencia, Spain
,
C Sanchez Montes
2   Hospital Universitari i Politècnic La Fe, Gastroenterology Department, Valencia, Spain
,
L Argüello Viudez
1   Hospital Universitari i Politècnic La Fe, Endoscopy Research Group. Gastroenterology Department, Valencia, Spain
,
V Pons Beltrán
1   Hospital Universitari i Politècnic La Fe, Endoscopy Research Group. Gastroenterology Department, Valencia, Spain
› Author Affiliations
Further Information

Publication History

Publication Date:
27 March 2018 (online)

Also available at
 

Aims:

To assess the accuracy of OD in a colorectal cancer (CRC) screening program, using NBI and the NICE classification, after following a non-guided self-learning program.

Methods:

Pilot prospective study in which a single endoscopist followed a learning process constituted by two phases. First, the NICE classification and a published set of still pictures describing OD of diminutive polyps was used. Second, OD was performed in 50 lesions detected during real-life colonoscopies with feed-back with the final pathological report. Second, OD was performed in consecutive CRC screening colonoscopies. Accuracy, sensitivity, specificity and predictive values of OD in 1 – 5 mm were calculated. The accomplishment of PIVI criteria for OD was also assessed. A minimum sample size of 277 polyps was calculated.

Results:

152 individuals were included [56 (36.8%) female; median age 61.1 (38 – 75)]. Overall, 522 lesions were detected. Diminutive (≤5 mm) lesions were the most prevalent [399 (76.4%)]. The majority of tdiminutive lesions were adenomas, 255 (63.9%). The examiner had a high confidence in OD in 347 lesions (87.0%).

Accuracy of PD: data on diagnostic accuracy for diminutive lesions are summarized in table 1.

Tab. 1

High confidence (95% IC)

Low confidence (95% IC)

Sensitivity

97.0 (95.6 – 98.3)

76.0 (72.6 – 79.4)

Specificity

74.3 (70.5 – 77.5)

46.0 (42.0 – 50.0)

Accuracy

89.0 (86.5 – 91.5)

60.0 (56.1 – 63,9)

PIVI criteria: The NPV for diminutive adenoma in the rectum was 0.92. Overall, 130 patients (85.5%) were suitable for scheduling the next follow up using OD.

The scheduled follow-up was correct in 120 (93.0%), 115 (89.1%) and 120 (93.0%) of patients following ESGE, EU and USA guidelines respectively.

Conclusions:

A non-guided self-administered learning program including real colonoscopies with pathology feedback is enough to learn OD and to fulfill the PIVI criteria when OD is applied in a CRC screening program.