Aims: This real-world observational clinical programme evaluated short and medium-term effects of intermittent flash glucose monitoring on HbA1c, glycaemic variability and lifestyle behavioural changes.Methods: Two first-generation Libre flash glucose monitoring sensors were provided 3-4 months apart with a food, activity diary, user evaluation survey and treatment modification after each sensor wear. T-tests were used to compare glucose variables within each sensor (week 1 vs. week 2) and between sensors (1st sensor vs. 2nd sensor). EasyGV software was used to calculate glycaemic variability.Results: From 42 type 1 diabetes and 120 type 2 diabetes participants, there was no statistically significant change in mean HbA1c for participants with type 1 diabetes at 3-4 months after the 1st sensor but there was a statistically significant HbA1c reduction for participants with type 2 diabetes [−4 mmol/mol (−0.4%), p = 0.008], despite no statistically significant differences in carbohydrate intake, exercise frequency and duration. Greater reduction was seen in those with baseline HbA1c> 86 mmol/mol (10%) in both type 1 [−12 mmol/mol (−1.1%), p = 0.009] and type 2 diabetes [−11 mmol/mol (−1.0%), p = 0.001). Both type 1 and type 2 diabetes showed improvements in Glucose Management Indicator and percentage time-above-range when comparing week 1 versus week 2 of the same sensor. Higher scan frequency resulted in improved glycaemic parameters and certain measures of glycaemic variability. The majority of participants (85%) agreed that flash glucose monitoring is a useful device but only 60% were keen to use it for daily monitoring. Conclusion:Constant feedback from flash glucose monitoring improves glycaemic parameters within the first week of wear. Intermittent use 3-4 months apart resulted in greater improvements for those with higher baseline HbA1c.

Objectives. We aimed to develop and validate a carbohydrate and insulin dosing knowledge quiz for adult Asian patients with diabetes mellitus.Methodology. A self-administered quiz was developed to test carbohydrate recognition; single food carbohydrate estimation; meal carbohydrate estimation and food label reading; and insulin dosing calculation for carbohydrate, blood glucose and for a meal in a multi-ethnic Asian population. The subjects' carbohydrate knowledge and insulin dosing ability were rated by the study dietitian and the subjects' primary physicians, respectively. We compared the quiz scores with the dietitians' and physician ratings and the subjects' HbA1c. Reliability of the quiz was tested by measuring internal consistency and split half reliability.Results. Seventy-five subjects completed the study. Median (inter-quartile range) quiz score was 71.9 (60.2 to 83.6)%. The quiz score was found to be correlated with the healthcare provider assessments (r=0.652, p<0.001) and the subjects' HbA1c (r=-0.375, p=0.001). Cronbach alpha was 0.897 and Guttman split half coefficient was 0.930.Conclusions. Our analysis suggested that this newly developed quiz had good reliability and validity for testing carbohydrate and insulin dosing knowledge in a group of Asian subjects with diabetes mellitus. This can be a useful screening tool in clinical practice.