In hospitalised patients, both extremes of glycaemia are associated with infections, mortality and increased costs. Guidelines generally recommend avoidance of hypoglycaemia and severe hyperglycaemia. We investigated if a patient’s capillary blood glucose level (BGL) on the first day predicts the development of a subsequent adverse glycaemic event (AGE) (defined as BGL <4.0 or >15.0mmol/L) during the admission.
The study was a prospective audit on 8 wards at a tertiary hospital over 2 months and is part of a baseline phase for a future intervention study (commencing May 2016). Consecutive inpatients with known diabetes or new-onset hyperglycaemia (random BGL ≥11.1mmol/L without history of diabetes) were recruited. Capillary BGLs were recorded using connectivity blood glucose meters (AMSL StatStrip), and commenced from time of admission to ward until discharge or day 14 for patients with prolonged hospital stay. Mean, minimum and maximum BGL of the first day (D1) were calculated using BGL from time of admission until end of the first full calendar day in hospital. Its association to a subsequent AGE (occurring from day 2 onwards) was determined using contingency analysis.
We recruited 342 patients totalling 1899 patient-days. AGEs occurred on 439 (23%) patient-days. Occurrence of AGE on D1 conferred a three-fold risk of a subsequent AGE during the admission (RR=3.1, 95%CI 2.2-4.3, p<0.001). Mean BGL on D1 was associated with the risk of subsequent AGE (χ2=27, p<0.001). D1 minimum BGL <5.0mmol/L and D1 maximum BGL >10.0mmol/L were associated with subsequent hypoglycaemia and severe hyperglycaemia respectively (p=0.001 for both). Of patients with all BGLs within range (5.0 – 10.0mmol/L) on D1, 21% still had a subsequent AGE.
Adverse glycaemic events in hospital are common. BGL measures on the first day of admission are associated with glycaemic extremes. The findings can guide future attempts to improve glycaemia and outcomes in diabetes inpatients.