• hypotension
• hypoglycaemia

Capillary blood glucose evaluation is routinely performed on patients presenting to the accident and emergency department. However, the limitations of this test are not widely known. We recently cared for a shocked patient who was hypoglycaemic (capillary glucose 1.3 mmol/l, venous laboratory glucose 2.3 mmol/l) on presentation. He was treated with repeated boluses of intravenous glucose and a single dose of intravenous glucagon (1 mg) as capillary blood samples remained hypoglycaemic. With continued resuscitation a further venous glucose sample revealed his formal blood sugar to be increased (30.8 mmol/l) while capillary levels were still in the hypoglycaemic range (1.8 mmol/l). We were unaware of the possibility of inaccuracy in this situation and discussion with colleagues revealed a similar lack of awareness.

Atkin et al¹ showed in a prospective study of hypotensive (systolic blood pressure <80) patients in the emergency department that 32% of patients were incorrectly diagnosed as hypoglycaemic by finger stick measurements. Indeed, on laboratory measurement of venous samples, two patients were hyperglycaemic. They recommended that venous blood samples measured with glucose reagent strips should be the preferred method of bedside blood glucose estimation in hypotensive patients as these results were comparable to laboratory values. The reason for the discrepancy between capillary blood glucose measurements and venous blood glucose measurements remains unclear. It has been proposed that, in the shocked patient, both peripheral vasoconstriction causing shunting of blood from the periphery and continued peripheral consumption lead to decreased capillary blood glucose concentrations.

While the risks of hypoglycaemia are widely appreciated, it is becoming increasingly recognised that hyperglycaemia is not desirable and may indeed worsen outcome.² The mechanism involved is uncertain but is probably related to increased cellular lactic acid production.³

Hypotension is frequently encountered in acutely ill patients and the limitations of a routinely used test need to be recognised and highlighted.