Abstract
INTRODUCTION: Pancreatic cancer is among the most aggressive malignancies, with a 5-year survival rate of 10%. Most patients present with advanced disease, limiting curative treatment options. Endoscopic ultrasound with fine-needle biopsy is the standard for diagnosis and staging. Although early access to endoscopic ultrasound (EUS) may enable timely systemic therapy and improve resectability, uncertainty remains regarding how delays to EUS affect surgical resection rates and overall survival, particularly in older adults. We aimed to identify factors associated with delayed EUS and to evaluate its impact on surgical resection and overall survival. METHODS: Using national Medicare claims (2011-2020), we conducted a retrospective cohort study of beneficiaries aged 66 years or older with newly diagnosed pancreatic cancer. The index date was the most recent claim for a pancreatic lesion or abnormal liver enzymes, serving as the indicator for EUS referral. Delay to EUS was defined as >30 days between the index date and the EUS procedure. Multivariable logistic regression identified sociodemographic and clinical factors associated with delayed EUS. Cox proportional hazards models estimated the associations between delayed EUS and 2 outcomes: (i) pancreatic surgical resection and (ii) all-cause mortality. RESULTS: Among 2,843 patients, 586 (20.6%) experienced a delay in EUS, 774 (27.2%) underwent surgery, and 1,591 (56.0%) died. Black patients were more likely to experience delay (adjusted odds ratio 1.65, 95%CI 1.09-2.51), whereas those with more comorbidities were less likely (adjusted odds ratio 0.95, 95%CI 0.90-0.99). Delayed EUS was associated with a lower likelihood of surgery (hazard ratio [HR] 0.73, 95%CI 0.61-0.88) but lower mortality (HR 0.58, 95%CI 0.50-0.66). Mortality increased with older age (HR 1.43, 95%CI 1.27-1.61) and comorbidity (HR 1.04, 95%CI 1.02-1.07). DISCUSSION: Timely EUS was associated with higher surgical resection rates, suggesting earlier access to curative treatment. Lower mortality among patients with delayed EUS possibly reflects disease severity confounding rather than benefit.