The effect of DAFNE education, continuous subcutaneous insulin infusion, or both in a population with type 1 diabetes...

To investigate the effect of DAFNE and continuous subcutaneous insulin infusion in clinical practice.

Within NHS Lothian, continuous subcutaneous insulin infusion started in 2004 and DAFNE education began in 2006. We extracted anonymized data from the national database for all those aged > 18 years with type 1 diabetes having a Dose Adjustment For Normal Eating course or continuous subcutaneous insulin infusion start date (n = 4617).

In total, 956 persons received DAFNE education, and 505 had received an insulin pump, 208 of whom had DAFNE education followed by insulin pump. Mean (SD) HbA_1c before DAFNE education was 68 (15) mmol/mol (8.4% [1.4%]) and 66 (13) mmol/mol (8.2% [1.2%]) before continuous subcutaneous insulin infusion. In the year following DAFNE education, the mean fall in within-person HbA_1c was 3.8 mmol/mol (95% CI 4.0 to 3.4; 0.3% [0.4% to 0.3%]). Those with the poorest control (HbA_1c ≥ 85 mmol/mol [9.9%]) experienced the largest decline (15.7 mmol/mol [1.4%]). Those in the lowest HbA_1c band at initiation (< 53 mmol/mmol [7.0%]) experienced a rise. In the year following continuous subcutaneous insulin infusion initiation there was a mean fall in within-person HbA_1c of 6.6 mmol/mol (6.8 to 6.4; 0.6% [0.6% to 0.6%]). In those with the poorest control (HbA_1c ≥ 85 mmol/mol [9.9%]), the mean fall in HbA_1c was 22.2 mmol/mol (23 to 21; 2.0% [2.1% to 1.9%]). Continuous subcutaneous insulin infusion effectiveness was not different with or without DAFNE education. The effects of both interventions were sustained over 5 years.

Both DAFNE education and insulin pump therapy had the greatest effect on HbA_1c in those with higher baseline values. There was little difference to attained HbA_1c when Dose Adjustment For Normal Eating education was introduced before insulin pump therapy.