An increased level of haemoglobin A1C predicts a poorer clinical outcome in patients with acute pancreatitis

OBJECTIVE: To compare the clinical features of acute pancreatitis (AP) in patients with and without diabetes. DESIGN, PATIENTS AND MEASUREMENTS: We retrospectively collected 318 patients with AP in two clinical centres from January 2009 to October 2010. Patients with a previous history of diabetes or with glycosylated haemoglobin A1c (HbA1c) higher than 6·5% were identified as having acute pancreatitis with diabetes (APD), while patients without a history of diabetes and in whom the HbA1c was not higher than 6·5% were considered as AP only. The clinical characteristics and prognosis data of these patients were analysed. Survival curves were plotted according to the Kaplan-Meier method. Cox proportional hazard regression was used to test the association between the clinical prognostic factors and mortality in patients with AP. RESULTS: In total, 318 patients with AP were enrolled. Among them, 40 had APD and 278 had AP; thus, in this study, the prevalence of diabetes in AP was 12·6% (40/318). Twenty-five per cent (10/40) of the APD cases were indentified using HbA1c. The mortality rate was significantly higher in the APD group (15·0%, 6/40) than that in the AP group (1·1%, 3/278). Survival curves showed that there was a significant survival difference between the APD group and AP group via the log-rank test. Multivariate Cox regression analysis showed that sex, age, diastolic blood pressure, body mass index (BMI) and C-peptide were significantly associated with mortality. Compared with AP patients, subjects with APD had significantly longer time from initial symptoms to admission [1·6 (95% CI: 0·5-3·2) vs 0·9 (95% CI: 0·1-2·2) days], older age of onset (57·2 ± 11·0 vs 44·3 ± 7·8 years), higher levels of glucose (13·9 ± 8·2 vs 7·3 ± 4·1 mm), higher levels of HbA1c [8·5 (95% CI: 6·6-11·4)%vs 5·9 (95% CI: 4·9-6·4)%], lower levels of C-peptide (0·882 ± 0·337 vs 2·621 ± 0·526 ng/ml) and longer duration of hospitalization (18·3 ± 4·6 vs 13·2 ± 5·1 day). Electrocardiograms showed that APD patients had a significantly higher risk of heart ischaemia than AP patients (22/40 vs 20/278). CONCLUSIONS: HbA1c may be a useful marker to identify unrecognised diabetes in patients with acute pancreatitis; this group of patients has a higher in hospital mortality.