ESPE Abstracts (2014) 82 P-D-1-3-83

DKA During Diabetes Therapy: Multinational Comparison with 59 191 Pediatric Patients from England, Wales, The United States, Austria and Germany

Justin Warnera, Julia Hermannb, Thomas Kapellenc, Sabine Hoferg, Stephanie Dubosee, Des Schatzf, Roy Becke, Claudia Schweigerh, David Maahsd & Reinhard Hollb


aChildren’s Hospital for Wales, on behalf of the National Paediatric Diabetes Audit and the Royal College of Paediatrics and Child Health, Cardiff, UK; bUniversity of Ulm, Ulm, Germany; cUniversity of Leipzig, Leipzig, Germany; dBarbara Davis Center for Childhood Diabetes, Denver, Colorado, USA; eJAEB Center, Tampa, Florida, USA; fUniversity of Florida, Florida, USA; gUniversity of Innsbruck, Innsbruck, Austria; hParacelsus Medical University, Salzburg, Austria


Background: DKA in children and adolescents with established type 1 diabetes (T1D) is a major problem with considerable cost to patients, families and health care systems. Many consider it as a quality of care indicator and a failure of relationship between the care provider and the family/patient. Considerable variability in rates are recognized. We analyzed multicenter registry and audit data from five countries with similarly advanced, yet differing, health care systems.

Objective and hypotheses: Data from 59 191 individuals <18 years with T1D from the T1D exchange (n=13 966, United States), the National Pediatric Diabetes Audit (n=18 963, England and Wales) and the DPV initiative (n=26 262, Austria and Germany) were pooled. DKA was defined as ≥1 hospitalization with a pH<7.3 during the prior year. Data were analyzed using multivariable logistic regression models for the whole population.

Results: The mean rate for DKA was 5.3% with differences amongst countries (6.2% US, 6.0% England, 4.5% Germany, 4.4% Wales, 3.3% Austria) which persisted after demographic adjustment (P<0.0001). Risk of DKA was highest in adolescents (14–18 years: 5.8%) compared to younger children (6–10 years: 3.4%; P<0.0001). DKA increased with longer duration of diabetes from 3.5% (<2 years) to 5.9% (≥2 years, P<0.0001). DKA was higher in girls compared to boys (OR=1.34 (1.24–1.44)). DKA was more prevalent in patients with ethnic minority status (OR=1.32 (1.21–1.43)). Additionally, DKA risk was lower in patients on insulin pumps (OR=0.88 (0.81–0.96)).

Conclusion: These multicenter data demonstrate important differences in DKA in childhood T1D across five nations. Benchmarking data such as these are important so countries can learn from each other to better understand where to target interventions in order to improve quality of care.