Aim SMART metres have taken paediatric diabetes management closer to home. Aim of our project was to enhance the learning with patients and their families toward home management thereby decreasing the need for hospital admissions and continuing support with the Diabetes MDT. SMART metre download review is a good way of analysing blood sugars targets, variability and control over a period of time.

Methods Patients and their families were taken through a process of ongoing learning to review and analyse SMART metre downloads and make appropriate changes to their insulin needs to prevent high and low sugars. The MDT had an oversight of the process to actively facilitate the learning. Data was collected from January 2014 to June 2014. A retrospective analysis was done on prospectively collected database of blood sugar downloads from SMART metres and near patient A1C tests.

Results Mean A1C for 100 downloads was 9.8 mmol/L that was comparable to a mean blood sugar of 9.6 mmol/L with a mean standard deviation of 4.7. However this correlation changed when the data was stratified based on Standard deviation (SD).

1. With SD < 2, the average A1C was 7.6 mmol/L compared to average mean blood sugar of 5.53mmol/L.

2. SD between 2– 4, co-related mean A1C of 8.7 mmol/L to average mean blood sugar of 7.9 mmol/L.

3. Surprisingly when SD was >4, the mean A1C–10 mmol/L and the mean average blood sugars–9.97 mmol/L were exactly the same.

4. This gap was widening the opposite way when the SD was >6 with A1C of 11.6 mmol/L compared to average mean blood sugar of 12.4 mmol/L.

This modality of reviewing and analysing results lead to better patient empowerment and care of their diabetes. Better control leads to better quality of life and comfort and confidence the children and their families with diabetes. There has been a 50% reduction of DKA and hypoglycaemia admissions on the ward with the use of SMART metres

Conclusion Simple SMART metres analysis are effective predictors for diabetes monitoring with average mean blood sugars which are well different to the nearer patient HbA1c and it bears a correlation between standard deviation of 4–6 with increasing gaps on both sides of the spectrum.