Trends in recorded capillary blood glucose and hypoglycaemia in hospitalised patients with diabetes.

AIMS: To utilise whole-system analysis of capillary glucose measurement results to examine trends in timing of glucose monitoring, and to investigate whether these timings are appropriate based on observed patterns of hypoglycaemia.
METHODS: Near-patient capillary blood glucose results from eight acute hospitals collected over 57 months were analysed. Analysis of frequency of measurement, and measurements in the hypoglycaemic (<4mmol/l) and severe hypoglycaemic (<2.5mol/l) range per time of day was made.
RESULTS: 3345241 capillary glucose measurements were analysed. 1657594 capillary blood glucose values were associated with 106624 admissions in those categorised as having diabetes. Large peaks in frequency of glucose measurements occurred before meals, with the highest frequency of capillary glucose measurement activity being seen pre-breakfast. Overnight, an increase in measurement activity was seen each hour. This pattern was mirrored by frequency of measured hypoglycaemia. 27968 admissions (26.2%) were associated with at least one hypoglycaemic measurement. A greater proportion of measurements were within the hypoglycaemic range overnight with 61.7% of all hypoglycaemia between 2100 and 0900h, with peak risk of measured capillary glucose being hypoglycaemic between 0300 and 0400h.
CONCLUSIONS: Hypoglycaemic is common with the greatest risk of hypoglycaemia overnight and a peak percentage of all readings taken being in the hypoglycaemic range between 0300 and 0400h. Measurement activity overnight was driven by routine, with patterns of proportion of measurements in the hypoglycaemic range indicating that there may be a significant burden of undiscovered hypoglycaemia in the patients not routinely checked overnight.