Efficacy and Safety of a Continuous Intravenous Insulin Protocol Modified for East Asians in Postoperative Glycemic Management Following Pancreatectomy.

Abstract

Introduction Glycemic control following pancreatectomy presents challenges, especially in patients with diabetes due to a lack of endogenous insulin, however, optimal management remains unclear. This study evaluated the efficacy and safety of a continuous intravenous insulin infusion protocol employed at our institution, in comparison with conventional glycemic control in patients with pancreatectomy. Materials and methods Sixty-one patients with preoperative glycosylated hemoglobin (HbA1c) of 6.5% or higher, on diabetes medications, or who underwent total pancreatectomy were included. Patients were categorized into three groups: insulin protocol (IP group, n = 24), subcutaneous injection (SI group, n = 15), and continuous intravenous insulin infusion based on the empirical control (EC group, n = 22). The primary outcomes were average blood glucose levels and the proportion of achievement within the target blood glucose range (140-180 mg/dl). Additionally, factors associated with the insulin dose in the IP group were analyzed. Results At predefined time points, the IP group achieved a significantly higher proportion of the target blood glucose range than the SI group (46.2% vs. 31.6%, p = 0.01), with no significant difference in average blood glucose levels (164.1 ± 41.8 vs. 169.1 ± 51.0 mg/dl, p = 0.50). During the 60-hour period following the initiation of frequent blood glucose measurements, the IP group demonstrated significantly reduced average blood glucose levels than the EC group (170.1 ± 56.0 vs. 175.5 ± 43.5 mg/dl, p <0.001), despite significantly longer measurement intervals (1.5 ± 0.7 vs. 1.2 ± 0.7 hours, p <0.001). However, there was no significant difference in the proportion of the target blood glucose range between the IP and EC groups (37.2% vs. 41.0%, p = 0.11). Aspartate transaminase and alanine transferase levels on postoperative day one were positively correlated with the average insulin dose in the IP group (both R = 0.45, p = 0.03). Conclusions This IP helped stabilize blood glucose levels compared to subcutaneous injections and improved glycemic control more effectively than empirically administered continuous intravenous insulin infusion. Postoperative elevations in liver enzymes may serve as predictors of increased insulin requirements.