BLUE LIGHT IMAGING FOR THE OPTICAL DIAGNOSIS OF SMALL COLORECTAL POLYPS: THE IMPACT OF A TRAINING INTERVENTION

Aims:

The advent of image enhanced endoscopic modalities have paved the way for better optical diagnosis of colorectal polyps. Blue Light Imaging (BLI) is a new technology that utilises powerful light emitting diode technology to enhance mucosal surface and vessel patterns. A specific BLI classification has recently been developed to enable better characterisation of colorectal polyps (BLI Adenoma Serrated International Classification – BASIC). The aim of our study was to investigate the diagnostic ability of BLI before and after training using this classification.

Methods:

BLI images from 45 polyps were shown to 10 endoscopists (5 experts with experience of advanced endoscopic imaging and 5 non-experts). They independently classified each of the images as adenoma or hyperplastic initially without any focused training on interpretation of BLI images. A face to face classroom training session was then delivered on BASIC and the endoscopists repeated the image classification exercise. The sensitivity, specificity, negative (NPV) and positive predictive values (PPV) for adenoma detection were calculated.

Results:

There was a significant improvement in sensitivity and NPV of adenoma detection as demonstrated in the table below (p < 0.05).

Tab. 1:

Diagnostic characteristics pre and post training

	Pre-training	Post-training
Sensitivity (95% confidence interval)	79.1 (73.3 – 84.2)%	95.7 (92.2 – 97.9)%
Specificity (95% confidence interval)	95.5 (91.8 – 97.8)%	91.8 (87.4 – 95.1)%
Positive predictive value (95% confidence interval)	94.8 (90.8 – 97.1)%	92.4 (88.7 – 95.0)%
Negative predictive value (95% confidence interval)	81.4 (77.3 – 84.9)%	95.3 (91.7 – 97.4)%

This improvement was mirrored in both expert and non-expert groups where sensitivity reached 97.4% in experts and 93.9% in non-experts, NPV reached 97.3% in experts and 93.2% in non-experts.

Conclusions:

The use of a bespoke BLI classification system with adequate training can significantly improve the sensitivity and NPV of adenoma detection thereby enabling the full potential of this novel imaging technology to be realised.