Endocrine Abstracts

Introduction: Effective management of diabetic ketoacidosis (DKA) improves clinical outcomes. We created an automated auditing system, Digital evaluation of ketosis and other diabetes emergencies (DEKODE), which identifies DKA episodes based on fixed-rate intravenous insulin infusion (FRIII) prescription.

Aim: We validated DEKODE for its ability to audit DKA management against manually collected data.

Methods: All episodes identified by DEKODE from September 2018 to August 2019 was compared with manually confirmed DKA episodes from the same duration. Duration of DKA, appropriateness of glucose and ketone measurements during entire DKA duration and fluids prescribed in the first 12 hours of diagnosis were compared between the two datasets. The difference between manual and automated data were analysed using Prism v6.0 (Graphpad Inc) and results are presented as mean and standard error of mean (SEM). The difference in frequencies of hypokalemia and hyperkalemia between manual and automated data was analysed by chi-square test.

Results: 150 episodes were identified by DEKODE. Of these, 147 had manually confirmed DKA. There was no significant difference in DKA duration between DEKODE and manual data (16.0 ± 1.0 hours; 17.5 ± 0.9 hours; P=ns) respectively. There was no difference in FRIII appropriateness (98.3% ±1.2%; 97.9% ± 1.1%; P=ns), glucose (98.5% ± 2.6%; 105.6% ± 2.5%; P=ns) and ketone measurements (43.3% ± 2.1%; 47.1% ± 2.2%; P=ns) between the two systems. DEKODE accurately predicted the frequency of hyperkalaemia (7/147; 6/150; P=ns) and hypokalaemia (9/147; 9/147; P=ns). However, DEKODE over-predicted proportion of fluids prescribed (96.9% ± 3.2%; 84.4% ± 3.1%; P=0.0047).

Conclusion: DEKODE reliably predicts DKA duration and management, which could help reduce time from data collection to analysis, thus providing real-time performance results.