How safe is endoscopic ultrasound-guided fine needle aspiration and biopsy (EUS-FNA/FNB) for the diagnosis of solid pancreatic masses?

Abstract

There is limited data on the prevalence of coexistent diabetes in acute pancreatitis and subsequent natural history in these patients.Using Pennsylvania Health Care Cost Containment Council data set, we identified 7399 unique White and Black Allegheny County, PA residents with first hospitalization for acute pancreatitis from 1996 to 2005. We evaluated the prevalence of coexistent diabetes, demographic and etiologic profile, severity of index hospitalization, and risk of readmission for acute or chronic pancreatitis during follow up (median 39 months) in this cohort.The prevalence of coexisting diabetes was 18% (1349/7399). When compared with non-diabetics, diabetics were significantly more likely to be older (63 vs. 56 yrs), male (OR 1.4, 95% CI 1.2–1.6), black (OR 2.4, 95% CI 2.1–2.7) and have non-alcoholic etiologies (biliary, OR 1.5, 95% CI 1.2–1.9; idiopathic, OR 2.0, 95% CI 1.7–2.5; metabolic, OR 5.2, 95% CI 4.0–6.7). While diabetic patients had a significantly longer length of stay (median 5 vs. 4 days, p < 0.05), their severity of acute pancreatitis (multivariable OR 1.18, 95% CI 0.94–1.48) or in-hospital mortality (1.9% each, p = 0.98) did not differ than non-diabetics. The overall risk of pancreatitis-related readmissions in diabetics was similar (33% each, p = 0.99), but their risk of subsequent admission for chronic pancreatitis (multivariable HR 0.65, 95% CI 0.44–0.97) was lower than non-diabetics.About 20% patients with acute pancreatitis have coexistent diabetes. Demographic and etiologic profile of diabetic patients with acute pancreatitis differs from non-diabetics. Diabetic status has limited effect on the severity of and natural history following acute pancreatitis.