INTRODUCTION: National studies on in-hospital pancreatic outcomes have focused on mortality. Non-fatal morbidity affects a greater proportion of patients.

METHODS: The Nationwide Inpatient Sample 1998-2006 was queried for discharges after pancreatectomy. Rates of major complications (myocardial infarction, aspiration pneumonia, pulmonary compromise, perforation, infection, deep vein thrombosis/pulmonary embolism, hemorrhage, or reopening of laparotomy) were assessed. Predictors of complication(s) were evaluated using logistic regression. Their independent effect on in-hospital mortality, length of stay, and discharge disposition was assessed.

RESULTS: Of 102,417 patient discharges, 22.7% experienced a complication. Complication rates did not decline significantly over time, while mortality rates did. Independent predictors of complications included age >or=75 [referent, 19-39; adjusted odds ratio (OR) 1.34, 95% confidence interval (CI) 1.2-1.5, p < 0.0001], total pancreatectomy (vs proximal, OR 1.29, 95%CI 1.1-1.5, p = 0.0025), and low hospital resection volume (vs high, OR 1.61, 95%CI 1.4-1.8, p < 0.0001). Complications were a significant independent predictor of death (OR 7.76, 95%CI 6.7-8.8, p < 0.0001), prolonged hospital stay (OR 6.94, 95%CI 6.2-7.7, p < 0.0001), and discharge to another facility (OR 0.28, 95%CI 0.26-0.3, p < 0.0001).

CONCLUSIONS: Despite improvements in mortality, complication rates remain substantial and largely unchanged. They predict in-hospital mortality, prolonged hospital stay, and delayed return to home. The impact on healthcare costs and quality of life deserves further study.