Sa1478 DOES OVERALL VOLUME OF ENDOSONOGRAPHER AFFECT OVERALL DIAGNOSTIC ACCURACY OF EUS-FNA OF SOLID PANCREATIC LESIONS: OUTCOMES FROM A LARGE HEALTH CARE SYSTEM COMPRISING OF BOTH ACADEMIC AND COMMUNITY PRACTICE

Abstract

Endoscopic ultrasound (EUS) guided sampling with fine needle aspiration (FNA) or fine needle biopsy (FNB) is the standard of care for diagnosis of solid pancreatic lesions. Over last decade, a large number of gastroenterologists have received dedicated training in EUS as well as ERCP. For ERCP, overall low volume is associated with more adverse events and negative patient outcomes. However, data regarding volume of endosonographer and overall clinical outcomes for patient is not available. Aim of this study was to examine impact of volume of endosonographer on overall adequacy and diagnostic accuracy of EUS guided FNA of pancreatic solid lesions (PSLs). EUS FNA is currently performed in 6 out of 14 hospitals of our health care system. Total 19 endosonographers performed over 3000 EUS between 2015 to 2019. From this prospective maintained database, we retrospectively reviewed EUS FNA cases for PSLs. One endosonographer performed 50 or more EUS FNA of PSL per year while other 18 endosongraphers performed less than 50 EUS FNA of PSL each per year. The collected data included patient demographics, PSL size, needle gauge, needle type, adequacy of the obtained sample, the percentage of diagnostic samples and accuracy of diagnosis. Mean age of the study population was 65+14. A total of 575 EUS-FNA of PSL were performed by all endosonographers. The high volume endosonographer (HVE) performed 242 EUS-FNA of PSL over a period of 4 years with more than 50 procedures per year. The low volume endosonographers (LVE) performed 333 EUS-FNA of PSL over the same period, each performing less than 50 procedures per year. There was no difference in mean size of PSL between both groups. Overall, LVE had significantly higher number 69/333 (20.7%) of inadequate samples compared to 40/242 (16.5%) inadequate samples for HVE (P <0.01). Similarly, overall diagnostic accuracy was much lower for LVE 244/333 (73.3%) compared to diagnostic accuracy of HVE 194/242 (80.2%) (P < 0.01). Overall low procedure volume of endosonographer is significantly associated with higher inadequate samples and lower diagnostic yield compared to high volume endosonographers.