R. Hanas, F. Lindgren and B. Lindblad (2009). A 2-yr national population study of pediatric ketoacidosis in Sweden: predisposing conditions and insulin pump use. Pediatric diabetes, 10(1), 33-7. https://doi.org/10.1111/j.1399-5448.2008.00441.x

PubMed 18761647

[Abstract]
The aim was to investigate triggering factors and insulin pump usage (continuous subcutaneous insulin infusion, CSII) at diabetic ketoacidosis (DKA). Data from 1999 and 2000 were collected retrospectively from Sweden. In 1999 and 2000, 7.4 and 11.0%, respectively, of children with diabetes used CSII. One hundred and forty-two episodes of DKA (pH < 7.30) were identified in 115 children (DKA at onset not included). Their hemoglobin A1c (HbA1c) was 10.1 +/- 2.0%, age 14.6 +/- 3.1 yr (range 1.5-19.9 yr), and diabetes duration 6.6 +/- 3.5 yr (range 0.4-17.7 yr). Fourteen persons (seven girls) had more than one episode of DKA. Reported causes of DKA were missed insulin doses (48.6%), gastroenteritis (14.1%), technical pump problems (12.7%), infection (13.4%), social problems (1.4%), unknown (5.6%), and not stated (4.2%). Alcohol was involved in eight episodes and drugs in one. Thirty of 115 patients (19 girls) used insulin pumps. The overall DKA incidence was 1.4/100 patient years in 1999 and 1.7/100 patient years in 2000. For insulin pump users, the DKA incidence was 3.2/100 patient years in 1999 and 3.6/100 patient years in 2000. HbA1c at DKA admission was lower for CSII users than patients who used injections (9.1 +/- 1.5 vs. 10.8 +/- 2.2%, p < 0.01), but pH and age did not differ. CSII had been used for 6 months (median) before the DKA episode. In conclusion, the DKA frequency in CSII users was approximately twice that of patients who used injections. Seventy-seven percent of the episodes occurred within 1 yr after CSII start. The high number of events reported to be caused by gastroenteritis is alarming because this may reflect a misinterpretation of DKA symptoms.