Point-of-Care Glucose

Point-of-care (POC) glucose testing measures glucose in a capillary (fingerstick), venous or arterial whole blood sample at the bedside with a handheld meter, using an enzymatic electrode strip (glucose oxidase or glucose dehydrogenase).

Results are ready in 5 to 10 seconds and are used for rapid diagnosis and treatment of hypoglycaemia and hyperglycaemia, insulin dosing, and glucose monitoring in ward, emergency, home and field settings.

Most meters report plasma equivalent glucose, with accuracy of within 15% of laboratory values (within 0.8 mmol/L when the laboratory value is under 5.5 mmol/L) in 95% of readings, and the best systems achieve 10% or less. Continuous glucose monitors measure interstitial glucose and give trend data but lag blood glucose by 5 to 15 minutes.

Diabetes affects over 590 million adults, and hypoglycaemia is the commonest acute complication of insulin and sulphonylurea therapy (severe hypoglycaemia in 1 to 3 episodes a year in insulin treated type 1 diabetes).

Hypoglycaemia in the neonate, malaria, sepsis, malnutrition, liver failure and alcohol intoxication can cause seizures and brain injury within minutes, and any person with altered consciousness, seizure or collapse needs an immediate glucose check. Hyperglycaemic emergencies (diabetic ketoacidosis, hyperosmolar hyperglycaemic state) are monitored hourly with POC glucose.

Indications. Altered consciousness, seizure, collapse, stroke symptoms, intoxication or unexplained confusion. Known diabetes on insulin or sulphonylureas, and intercurrent illness. Diabetic ketoacidosis and hyperosmolar states (hourly during treatment). Neonates at risk of hypoglycaemia (small for gestational age, preterm, infant of a diabetic mother).

Severe malaria, sepsis, malnutrition, liver failure. Pregnancy with gestational diabetes. Before and after exercise in type 1 diabetes. Parenteral nutrition and corticosteroid therapy. Screening in community settings where laboratory tests are not available. Not for diagnosis of diabetes (laboratory plasma glucose or HbA1c confirm).

Specimen. Capillary blood from the side of a fingertip (after washing and drying hands; alcohol swabs must evaporate; do not squeeze or milk the finger; discard the first drop if the finger is contaminated), heel in neonates, or venous or arterial whole blood. Test strips must be in date, stored sealed at room temperature, and matched to the meter code. Apply a sufficient drop to the strip. Record the time, relation to meals and insulin, and the patient's identity.

Reference values. Fasting 3.9 to 5.5 mmol/L (70 to 100 mg/dL). Random 3.9 to 7.8 mmol/L. Two hours after a meal under 7.8 mmol/L. Neonates 2.6 mmol/L or more after the first hours of life. Targets in diabetes are individualised: fasting 4 to 7 mmol/L and under 10 mmol/L after meals in adults with type 1 diabetes, with wider ranges in older frail patients and hospital inpatients (6 to 10 mmol/L, up to 12 mmol/L).

Interpretation.

• Level 1 hypoglycaemia: under 3.9 mmol/L (70 mg/dL), an alert value to treat. Level 2: under 3.0 mmol/L (54 mg/dL), clinically significant. Level 3: severe hypoglycaemia with altered mental or physical state needing help from another person, at any glucose.
• Whipple triad confirms hypoglycaemia: symptoms, low measured glucose, and relief of symptoms when glucose is raised.
• Hyperglycaemia: random glucose of 11.1 mmol/L (200 mg/dL) or more with symptoms is diabetic range; fasting 7.0 mmol/L or more on laboratory plasma glucose, or 2-hour glucose 11.1 mmol/L or more on oral glucose tolerance test, or HbA1c 6.5% (48 mmol/mol) or more is diabetes, confirmed by repeat testing when asymptomatic.
• Glucose over 14 mmol/L with ketonaemia (beta hydroxybutyrate 3.0 mmol/L or more) or ketonuria and acidosis: diabetic ketoacidosis. Glucose over 30 mmol/L with high osmolality and little ketosis: hyperosmolar hyperglycaemic state.
• Stress hyperglycaemia in acute illness (sepsis, stroke, myocardial infarction, steroid use) is common, and persistent hyperglycaemia needs follow up for diabetes.
• Nocturnal and early morning pattern checks separate the Somogyi effect, dawn phenomenon and insufficient basal insulin.
• Self-monitoring records, time in range (3.9 to 10.0 mmol/L, target over 70%) and glucose management indicator from continuous monitors complement HbA1c.
• A low reading unrelated to symptoms is confirmed by repeat testing or laboratory glucose, but treatment is not delayed when symptoms are present.

Critical values and findings needing immediate action.

• Glucose under 3.0 mmol/L, or any glucose with symptoms (confusion, seizure, coma): treat at once. Conscious patient: 15 to 20 g of fast acting carbohydrate (glucose tablets or juice) and repeat in 15 minutes. Unconscious patient: intravenous 10 to 20% glucose (for example, 75 to 150 mL of 20% glucose in adults, 2 mL/kg of 10% glucose in children and neonates) or intramuscular glucagon 1 mg, then a long acting carbohydrate.
• Sulphonylurea or long acting insulin hypoglycaemia: prolonged observation (24 hours or more) and further treatment with octreotide for recurrent episodes.
• Glucose over 15 mmol/L with vomiting, abdominal pain, rapid breathing or drowsiness: check ketones and venous blood gas for diabetic ketoacidosis.
• Glucose over 30 mmol/L with confusion and dehydration: hyperosmolar state, fluids and insulin, with thromboprophylaxis.
• Neonate with glucose under 2.6 mmol/L: feeding or intravenous glucose, and investigation of cause.
• Any patient with a seizure, coma or sudden neurological deficit: check glucose before further interventions.

Interferences and pitfalls.

• Meters are less accurate in the extremes (under 3.0 and over 20 mmol/L), in shock, severe dehydration, hypotension and peripheral vasoconstriction (capillary samples become unreliable), and in the critically ill, where arterial or venous samples on a blood gas analyser are preferred.
• Haematocrit: high haematocrit (neonates, polycythaemia) gives lower readings, and anaemia gives higher readings on many meters.
• Interfering substances: maltose, galactose and xylose (glucose dehydrogenase pyrroloquinoline quinone strips give false highs during icodextrin dialysis, immunoglobulin infusions and some enzyme tests), paracetamol, ascorbic acid, uric acid and high dose vitamin C, dopamine, and mannitol. Strip chemistry must be known.
• Oxygen: glucose oxidase strips are affected by high oxygen tension (oxygen therapy) and hypoxia.
• Contamination of the fingertip with sugary food, fruit or cosmetics causes false highs; alcohol wipes not dried cause false lows.
• Strips damaged by heat, humidity, expired dates or open vials give errors; control solution is run at set intervals.
• Meters are not used for the diagnosis of diabetes, and neonatal hypoglycaemia needs laboratory confirmation; low neonatal readings are confirmed on a blood gas analyser or laboratory sample.
• Continuous monitors lag by 5 to 15 minutes, and need fingerstick confirmation when symptoms disagree with readings or during rapid change.
• Blood contamination risks: single use lancets and strict hygiene prevent transmission of hepatitis B and other infections.
• User technique, training and recording errors matter, and competency checks and quality assurance are needed in ward use.

Special groups.

• Neonates and children: hypoglycaemia thresholds vary by age; heel prick samples are used in neonates.
• Pregnancy: tight targets (fasting under 5.3 mmol/L, 1 hour after meal under 7.8 mmol/L, 2 hours under 6.4 mmol/L), because of foetal risk, and frequent testing in insulin therapy.
• Older adults: avoid hypoglycaemia, which causes falls, cognitive harm and cardiac events, with relaxed targets.
• Renal impairment and dialysis: altered insulin clearance, and risk of interference.
• Critically ill patients: arterial or venous samples are used for accuracy, with fixed local insulin infusion schedules.
• Regions: POC glucose meters allow rapid assessment in malaria, malnutrition, trauma and community diabetes screening in areas without laboratories, where quality control and strip supply limit reliability.

Follow up tests. Laboratory plasma glucose to confirm unexpected values, HbA1c, ketones (blood beta hydroxybutyrate), venous blood gas with electrolytes and lactate, urea, creatinine and electrolytes, insulin and C-peptide with sulphonylurea screen in unexplained hypoglycaemia during a low glucose episode, cortisol and growth hormone in hypoglycaemia in children, liver function,

thyroid function, urine ketones and albumin to creatinine ratio, lipid profile, retinal screening and foot assessment in diabetes, and review of meter technique and continuous glucose monitoring data.