HPB P43 Pancreatic Anastomosis Audit (PARANOIA): Initial experience of an international POPF registry

Abstract

Abstract Background Post-operative pancreatic fistula (POPF) remains the primary determinant of poor outcome following pancreatoduodenectomy (PD) Risk factors are well understood but there is no apparent reduction in incidence. RCT’s are methodologically flawed to investigate POPF. PARANOIA is a registry with two novel outcomes. 1) Surgeons-specific risk-adjusted outcomes: Surgeons will observe in real time how changes to surgical technique impact upon outcomes. By having access to the registry data surgeons can explore technical factors of pancreas surgery that are associated with high or low risk of pancreatic fistula. 2)The registry will serve as a platform for international prospective studies. Methods Core dataset consists of common risk factors found within scores (aFRS, FRS, Roberts’) along with anastomotic technique and adjuncts. Initial prospective studies test topical clinical questions through embedded parallel studies. Using a stepwise approach, the registry and user network will be tested with increasingly complex study designs to determine the functionality of the registry as a novel research tool. Surgeons will have their risk-adjusted outcomes as a CUSUM plot which they can compare against anonymised surgeons by different variables. Results 232 registered surgeons across 56 countries with 776 cases. Majority of surgeons are performing open surgery (open 208 (89.7%) vs laparoscopic (11 (5.2%) vs robotic 11 (5.2%)). End-to-side was the commonest technique (88.8%) with invagination (69%) being used more than duct-to-mucosa (31%). 43.5% of surgeons used an internal stent routinely and almost all used drains (97.8%). 13% used pharmaceutical sealants whereas 28.4% used an autologous barrier. 44% used a somatosatin analogue (SSA). 70.5% reported no POPF whereas 8.4% had a CR-POPF rate. A reduced CR-POPF was associated with use of stent (p=0.008), autologous barrier (p=0.012) and SSA (p=0.049). Conclusions The PARANOIA registry has shown rapid recruitment of surgeons and data entry in pancreatic surgery can not only shorten the timeframe from study conception to dissemination of results but also provide real world data and provide comparison of outcomes between and within geographical regions, a key factor given that some patient and operative variables relevant to POPF vary across the globe. Registry data and the risk-adjusted CUSUMS can be used by surgeons to benchmark themselves against their peers and consider how modifications to their operative technique or perioperative care may affect their outcomes.