THE OOPS STUDY: THE OOSTENDE POLYP SIZING STUDY

Aims Polyp size is one of the defining characteristics after polypectomy to decide the surveillance recommendations. In current practice, this is often based on endoscopic estimation.

We wanted to investigate the accuracy of endoscopic sizing, define its impact on surveillance recommendations and explore teachability.

Methods A monocentric, prospective study was performed between April and October 2018 and February and November 2019. During the first period endoscopists were confronted with their performance, during the second they were blinded. Endoscopists sized polyps by visual estimation without support of devices. This was then compared to the gold standard ex-vivo prefixation and postfixation measurements. To eliminate the risk of electrocautery induced polyp shrinkage, only pedunculated polyps were included. Other factors that influence surveillance recommendation were collected.

The difference between estimation and prefixation was compared between the different inclusion periods.

Normality was tested with Shapiro-Wilk and continuous variables were compared with a paired sample T-test or Mann-Whitney U.

Results In total 128 polyps were included, of which 72 in the initial and 56 in the subsequent inclusion period. The endoscopic sizing (12.98±5.32mm) differed statistically from the prefixation (11.53±4.96mm; p< 0.001) and postfixation (11.75±4.81mm; p< 0.001) measurements. Overestimation (62.5%) was seen more often than underestimation (21.1%). There was no statistical significant difference between prefixation and postfixation measurements (p=0.297).

An inappropriate surveillance recommendation was given in 22.7% of the cases, of which 93.1% were due to overestimation. After considering other factors that define the surveillance recommendation, still 13.3% were erroneous.

Concerning teachability, no statistically significant improvement in estimation error between the initial (1.67±2.88mm) and subsequent (1.39±2.76mm) inclusion period could be established (p=0.753).

Conclusions Endoscopic polyp sizing is prone to overestimation resulting in inappropriate surveillance recommendations. There only seems to be limited teachability towards endoscopic sizing without the use of devices.

Until the advent of new measuring devices, we propose using ex-vivo measurements in pedunculated polyps to establish post-polypectomy surveillance recommendations.