In-hospital outcomes of patients with pancreatic cancer and venous thromboembolism

e16317 Background: Pancreatic cancer accounts for about 3.2% of all new cancers in the US. It has dismal 5-year relative survival of 11.5% and accounts for 8.2% of all cancer deaths. Studies have shown a strong association of pancreatic cancer with venous thromboembolism (VTE). However, data surrounding outcomes associated with VTE in this population is limited. We aim to estimate the effect of VTE on mortality and healthcare utilization in this population. Methods: We queried the National Inpatient Sample (NIS) 2017-2019 for all adult patients with a diagnosis of pancreatic cancer. The cohort was further classified by the presence of VTE (either acute deep vein thrombosis (DVT) or acute pulmonary embolism (PE)). The primary outcome of interest was in-hospital mortality, and secondary outcomes included hospitalization cost and length of stay, reflecting resource utilization. Multivariable regression analysis was performed to examine the association. Results: Over a period of 3 years, 392,744 admissions were identified with a diagnosis of pancreatic cancer. The mean age was 67 years, and 48% were females. The prevalence of acute DVT and acute PE was 5.0% and 5.8% respectively. In-hospital mortality was roughly 12% in the study cohort who had either DVT or PE. The odds of all-cause mortality were higher in pancreatic cancer admissions complicated by DVT (OR 1.59, 95% CI: 1.34-1.70, p<0.001) or PE (OR 1.39, 95 %CI: 1.23-1.58, p<0.001). The presence of VTE was also associated with higher hospital costs ($17,351 (14,256-20,446, p<0.001) for DVT, and $5,710 (2,679-8,740, p<0.001) for PE and longer hospital stay (7.2 vs 5.7, p<0.001) for DVT and (6.6 vs 5.7, p<0.001) for PE. Conclusions: Our study suggests that the presence of VTE portends poor outcomes in patients with pancreatic cancer. They also increase healthcare expenditure and lead to a longer hospital stay. Current guidelines (2020) recommend against offering thromboprophylaxis to patients with any cancer. Given the worse outcomes, this study reflects the need to formulate risk assessment models for the role of primary thromboprophylaxis in pancreatic cancer.