European Stroke Organisation (ESO) guidelines on glycaemia management in acute stroke

Abstract

Background: Hyperglycaemia is a frequent complication in acute stroke that has been shown to be independently associated with larger infarct size, haematoma growth, poor clinical outcome and mortality. This Guideline Document presents the European Stroke Organisation (ESO) Guidelines for the management of blood glucose levels in patients with acute ischemic or haemorrhagic stroke.

Methods: The working group identified related questions and developed its recommendations based on evidence from randomised controlled trials following the standard operating procedure of the ESO. This Guideline Document was reviewed and approved by the European Stroke Organisation Guidelines Committee and the European Stroke Organisation Executive Committee.

Results: We found low-quality evidence from clinical trials in ischemic or haemorrhagic stroke exploring the use of intravenous insulin aimed to achieve a tight glycaemic control with different glucose level targets and several other sources of heterogeneity. None of these trials neither the meta-analysis of them have demonstrated any significant benefit of tight glycaemic control with intravenous insulin in acute ischemic or haemorrhagic stroke patients on functional outcome or in survival and they have shown an increased risk for hypoglycaemia.

Conclusions: We suggest against the routine use of tight glycaemic control with intravenous insulin as a means to improve outcomes. The currently available data about the management of glycaemia in patients with acute stroke are limited and the strengths of the recommendations are therefore weak. Nevertheless, this does not prevent that hyperglycaemia in acute stroke patients could be treated as any other hospitalised patient.

Keywords: Acute stroke; European Stroke Organisation; GRADE; guidelines; hyperglycaemia.

Figure 1.

Flow diagram of the systematic review search. *One manuscript excluded as this was only published as conference proceedings. One manuscript excluded due to nonconfirmation of stroke prior randomisation and another one due to active treatment with IV insulin in the control group. Two manuscripts^, excluded due to the impossibility to obtain separate data for ischemic and haemorrhagic stroke and data collection not currently available after contacting the authors.

Figure 2.

Forest plot analysis for ischemic stroke. (a) Good functional outcome at the end of follow-up, (b) survival at the end of follow-up, (c) hypoglycaemic events (<3–4 mmol/L) and (d) symptomatic hypoglycaemic events.

Figure 3.

Forest plot analysis for ischemic stroke patients with hyperglycaemia at baseline. (a) Good functional outcome at the end of follow-up, (b) survival at the end of follow-up, (c) hypoglycaemic events (<3–4 mmol/L) and (d) symptomatic hypoglycaemic events.