The impact of hypoglycaemia on patients admitted to hospital with medical emergencies

Aims To quantify the frequency of biochemical hypoglycaemia in acutely unwell patients in the medical assessment unit and relate this to their subsequent outcomes. Methods A retrospective audit was conducted on all emergency medical patients attending the medical assessment unit between November 2010 and April 2011. Capillary blood glucose measurements were obtained and electronically stored for all patients. Admission details, presence of diabetes, type of diabetes and treatment, length of stay in hospital and death in hospital were obtained from the hospital clinical coding data and electronic discharge summary. The incidence of hypoglycaemia in patients with and without diabetes was quantified. The mean age, length of stay and percentage of death in hospital were compared between groups with and without hypoglycaemia. Results One hundred and thirty-eight (9.5%) patients with diabetes and 70 (2.7%) patients without diabetes had an episode of hypoglycaemia in the medical assessment unit. Patients with diabetes and hypoglycaemia on admission had a significantly longer length of stay (mean +/- sd) (10.3 +/- 11.2 vs. 7.3 +/- 9.5days, P=0.001) and higher rate of hospital mortality (14.5 vs. 5.2%, P<0.001) compared with those without hypoglycaemia. Patients without diabetes with hypoglycaemia had a longer length of stay (mean +/- sd) (9.1 +/- 10.5 vs. 6.7 +/- 9.9days, P=0.05) and a higher rate of hospital mortality (24.3 vs. 5.4%, P<0.001) compared with those without hypoglycaemia. Conclusion Hypoglycaemia is associated with an increased length of stay in hospital and an increased in-hospital mortality rate. Hypoglycaemia may have contributed to the poorer outcome, but would also appear to be a marker of disease severity in unwell patients, especially patients with sepsis.