Aims/hypothesis: We aimed to describe the frequency and degree of diabetic ketoacidosis in children across Europe at the time of diagnosis of Type I (insulin-dependent) diabetes mellitus and to determine if factors such as age and geographical region contribute to the risk of diabetic ketoacidosis.
Methods: The study was part of the EURODIAB project. A total of 24 centres, covering a population at risk of more than 15 million children below 15 years of age, recruited 1,260 children at the time of clinical diagnosis.
Results: Polyuria, by far the most frequent symptom, was observed in 96% of the children. In only 25% of the children was the duration of symptoms less than 2 weeks and this proportion was larger in the under 5 year age-group (37 vs 22%; p < 0. 001). Of the 11 centres that recorded diabetic ketoacidosis status, the overall proportion with diabetic ketoacidosis (pH < 7.3) was 40% (95%-CI: 36-44%) in at least 90 % of cases. After stratification by centre, the odds ratio for diabetic ketoacidosis in the under 5 age-group was 1.02 (95%-CI:0.69-1.49) relative to the older children. There was significant variation between the 11 centres in the frequency of diabetic ketoacidosis which ranged from 26 to 67% (p = 0.002). An inverse correlation between the frequency of diabetic ketoacidosis and the background incidence rate was found in these centres (Spearman's rank correlation, rs = -0.715;p = 0.012).
Conclusion/interpretation: Rising standards of medical information and greater awareness concurrent with an overall increase in incidence could have resulted in changes in the clinical presentation at onset of Type I childhood diabetes in Europe.