A retrospective study on comparison of clinical characteristics and outcomes of diabetic ketoacidosis patients with and without acute pancreatitis

Abstract

The co-existence of diabetic ketoacidosis (DKA) with acute pancreatitis (AP) is associated with unfavorable clinical outcomes. However, diagnosing AP in DKA patients is challenging and often missed due to overlapping symptoms. The aim of this retrospective observational study was to compare the clinical characteristics and outcomes of patients with concomitant DKA and AP or DKA alone. Data of patients with DKA admitted between January 2015 to August 2021 to four hospitals in Qatar was extracted from the electronic health record (Cerner). American Diabetes Association criteria and Atlanta criteria were used for DKA and AP diagnosis, respectively. Independent T-test or Mann–Whitney U test was used to analyze continuous variables, whereas categorical variables were analyzed via Chi-square or Fischer exact tests as appropriate. Univariate and multivariate logistic regression models were generated to assess the correlations. A p-value of < 0.05 was considered statistically significant. Of 936 patients with DKA, 84 (9.0%) had coexisting AP. AP was most common in the Asian race (66%, p < 0.001). Patients with DKA and AP were older, had higher admission anion-gap, white cell count, hemoglobin (hb), neutrophil/lymphocyte ratio, urea, creatinine, maximum blood glucose during the episode, total cholesterol and triglyceride level (TGL) (p < 0.05). They had a lower admission venous pH and bicarbonate at 6 h. Patients in the DKA with AP group also had a longer length of stay (LOS), DKA duration and a higher rate of ICU admission (p-values ≤ 0.001). In-hospital mortality, 3-month all-cause readmission, 6-month and 12-month DKA recurrence did not differ between the two groups. Univariate logistic regression analysis showed age, Asian ethnicity, male gender, T2D, admission WBC count, hb, urea, creatinine, potassium, venous pH, bicarbonate, anion gap, total cholesterol, TGL and LDL level were significantly associated with the development of DKA with AP (p < 0.05). In multivariate logistic regression analysis, age and total cholesterol level were associated with concomitant DKA and AP (p < 0.05). Patients with concomitant DKA and AP have more severe derangement in markers of DKA severity, inflammation, kidney injury and metabolic profile, along with a longer DKA duration, LOS and requirement for ICU support compared to DKA patients without AP. This highlights the clinical significance of diagnosing the co-existence of DKA with AP, as the combination results in significantly worse clinical outcomes and greater healthcare utilization than in patients with only DKA.