Admission blood glucose helps predict 1 year, but not 2 years, mortality in an unselected cohort of acute general medical admissions.

Abstract

We previously showed that hyperglycaemia in newly hospitalised medical inpatients is associated with longer length of hospital stay, higher 28-day readmission rates and increased 28-day mortality. We aimed to assess whether a single blood glucose measurement taken at the time of admission could help to predict 1 and 2 years mortality. We retrospectively reviewed data from all 1502 patients admitted to our Acute Medical Unit during February 2010. By using a blood glucose range of 6.5-7.0 mmol/l as the comparator, an admission blood glucose between 9.1 and 20 mmol/l was associated with an increased risk of death at 1 year (p < 0.05). In addition, those people with admission glucose readings of < 6.5 mmol/l showed a strong trend towards a higher mortality (p = 0.053) at 1 year. Thus admission blood glucose can be used to help predict the risk of 1 year mortality in an unselected cohort of general medical admissions.