Hyperglycaemia in acute stroke--to treat or not to treat.

Diabetes is common amongst patients with stroke and is associated with poorer outcome. Post-stroke hyperglycaemia is also recognised in up to half of the patients, and is independently associated with adverse sequelae: both increased mortality and poorer functional outcomes. Neither the aetiology nor the pathophysiology of such hyperglycaemia is fully understood. Both direct neurological toxicity and systemic consequences are postulated to occur. A distinction between occult diabetes and non-diabetic hyperglycaemia seems important as prognosis and effect of intervention differ in these two groups. The optimal management of the milder forms of hyperglycaemia associated with acute stroke is unknown. Randomised trial data remain limited but presently offer no strong support for aggressive intervention in stroke, though in other critical illness settings tight control of blood sugar appears beneficial. Studies based in coronary care and high dependency units have shown a possible beneficial effect of insulin, but evidence for intervention in acute stroke is at best limited. However, if glucose management is to be undertaken, this should be instituted while there is still salvageable tissue and the glucose reduction must be substantial. Intravenous insulin may be more effective than glucose-potassium-insulin infusion. Both interventions carry a risk of hypoglycaemia and any proposed intervention must balance safety, convenience and glycaemic control. Until further trial data are available, consensus guidelines may be followed, which are generally conservative for blood glucose levels below 10 mM (180 mg/dl). (c) 2009 S. Karger AG, Basel.