ESPE Abstracts

Background: The use of insulin pumps is rapidly spreading within the paediatric T1D community. A few small studies have promoted pump usage, while large registries have shown almost no advantage of this treatment.

Objectives: Compare the results of treatment with insulin pump (‘Pump’) with insulin injections (‘Inj’) in a large cohort of children with T1D in field conditions.

Patients: We studied 3137 children of the ISIS-Diab cohort with ≥1 year of T1D duration recruited by 99 diabetes centers of various size covering all French regions. HbA1c, severe hypoglycemia (SH), ketoacidosis (DKA), and weight were compared between ‘Pump’ (n=694) and ‘Inj’ (n=2443) groups of children over their last 6 months of treatment.

Results: The ratio of ‘Pump’/‘Inj’ varied from 0 to 65% across centers. ‘Pump’ children were slightly younger (9.2±3.9 years vs 10.6±3.4 years for ‘Inj’, P=2.10⁻⁶), and had comparable socio-educational level than ‘Inj’. Mean HbA1c was 7.8±1.0% in ‘Pump’ and 7.9±1.2% in ‘Inj’ (NS). SH occurred in 3.5% of ‘Pump’ and 5.4% of ‘Inj’ (P=0.053). DKA occurred in 2.7% of ‘Pump’ and 2.7% of ‘Inj’. Overweight was present in 10.7% of ‘Pump’ and 10.4% of ‘Inj’. Insulin dosage was 0.82±0.28 U/k.d in ‘Pump’ and 0.95±0.29 U/k.d in ‘Inj’ (P<2.10⁻¹⁶). Large expert centers had the same pump results than small centers (HbA1c 7.8±1% vs 7.8±1.0%).

Conclusion: In field conditions, insulin pump shows no clear superiority. The choice of this costly and more demanding mode of treatment should thus be balanced at the individual and public health level.