Introduction

Admission hypoglycaemia is associated with increased hospital mortality.  However, it is not known if hypoglycaemia directly contributes to death or is merely a marker of increased severity of illness.

Methods    

We previously published a prospective observational study where blood glucose level (BGL) was measured routinely on 6187 consecutive patients admitted through Emergency.¹  In this sub-study, the hypoglycaemic cohort (BGL<4 mmol/L, n=110) was reviewed to identify contributing factors, recognition and treatment of the hypoglycaemia, and relationship to mortality.  Contributing factors examined included insulin or sulphonylurea therapy, malnutrition, acute or chronic heart failure, renal failure, liver failure, malignancy, cortisol deficiency, acute alcohol excess, pregnancy and sepsis.

Results

Hypoglycaemia was associated with increased mortality (p=0.008), compared to the normoglycaemic group (BGL 4-7.9mmol/L). Median BGL of the hypoglycaemic cohort was 3.5 mmol/L (range 1.1-3.9) and age 63.5 years (range 17-84).  Thirty-eight patients (35%) had known diabetes, and most were treated with insulin or sulphonylurea therapy (33/38). Ten patients died (9.1%), 6 had diabetes.

Most patients (70%) had ≥1 identifiable contributing factor for hypoglycaemia, 43% had ≥2.  Patients with diabetes had more contributing factors, excluding diabetes treatment (p=0.001).  Recognition and follow-up glucose testing were more common in patients with diabetes (p<0.001), but even then 27% went unrecognised and untreated.  Patients with unrecognised hypoglycaemia who died, had multiple contributing factors.  Death was more likely with increasing number of contributing factors (p<0.001), but degree of hypoglycaemia was not associated (p=0.472), and no death was directly related to the hypoglycaemia.  No patients without any identifiable contributing factor died, and in those who were recognised but not treated, hypoglycaemia was mild. 

Conclusion

Admission hypoglycaemia is likely to represent a marker of underlying disease severity, rather than a direct cause of death.  “Stress Hypoglycaemia” may be an under-recognised phenomenon that identifies patients on admission at increased risk of death.