167-OR: Significant Differences in DKA Outcomes across Hospitals Despite Same Guidelines Highlights the Need for Better Implementation Strategies—Results from DEKODE Initiative

Abstract

Aim: To measure differences in DKA care and outcomes between hospitals having same guidelines. Method: The data was gathered as part of the DEKODE (Digital Evaluation of Ketosis and Other Diabetes-related Emergencies) initiative from October 2021 to September 2022 in six hospitals (anonymised as A to F) in the UK which has guidelines based on Joint British Diabetes Societies-Inpatient (JBDS-IP) recommendations. We collected data on adherence to Fixed Rate Intravenous Insulin Infusion (FRIII), fluids administration, glucose monitoring, and ketones monitoring. Pseudonymised data on dates and times of admission, discharge, DKA diagnosis and resolution was collected. Independent-Samples Kruskal-Wallis test was performed to assess differences in various parameters of DKA care and outcomes. Results: A total of 442 DKA episodes were analysed. There were differences between hospitals in fluids administration (median: A-92.1%, B-101.3%, C-89.7%, D-76.0%, E-98.3%, F-115.5%; p<0.001), glucose monitoring (median: A-103.1%, B-86.1%, C-88.9%, D-106.4%, E-99.2%, F-95.7%; p<0.001), and ketone monitoring (median: A-71.5%, B-85.1%, C-70.0%, D-59.3%, E-65.0%, F-85.7%; p<0.001). There was no significant difference in FRIII (median: A- 99.5%, B- 100.0%, C- 100.0%, D- 100.0%, E- 93.9%, F- 99.8%; p=0.064). We found inter-hospital differences in DKA duration (median in hours; A-13.3, B-13.4, C-18.5, D-15.1, E-17.1, F-16.3; p=0.002), and length of stay (median in days; A-2.1, B-2.5, C-3.5, D-3.6, E-2.5, F-3.0; p<0.001). Conclusion: Despite similar guidelines, significant variations in DKA outcomes exist between hospitals. Variations in various components of DKA highlight there may be differences in guidelines implementation. A centralised data collection system can identify and share best practices between centres, which may help minimise inter-hospital variations and facilitate uniform patient care. Disclosure P.Kempegowda: None. L.Rengarajan: None. P.Narendran: Advisory Panel; Omnipod, Speaker's Bureau; Abbott Diabetes, Lilly Diabetes. Dekode team: n/a.