Management of Blood Glucose in ICU

Blood glucose management

Synonyms

Glycemic control, hyperglycemia, hypoglycemia, intensive insulin therapy, conventional insulin therapy

Related conditions

Insulin therapy, blood glucose measurement, enteral and parenteral nutrition, hepatic failure, endocrine disorders, insulinoma

1. Description of the problem

What every clinician needs to know

Hyperglycemia (defined as blood glucose concentration [BG] >110mg/dL, [>6.1 mmol/L]) occurs in more than 90% of critically ill patients. Both the occurrence and degree of hyperglycemia are independently associated with increased morbidity and mortality. Both severe hypoglycemia ([BG] <41mg/dL), which occurs in 3-5% of critically ill patients, and mild to moderate hypoglycemia ([BG] <71mg/dL), which occurs in up to 25%, are independently associated with increased mortality. The incidence of hypoglycemia is increased by targeting a lower blood glucose concentration. Increased [BG] variability is also associated with increased mortality.

Clinical features

Unless extreme, hyperglycemia in the context of critical illness is generally asymptomatic and gives rise to no specific clinical signs. Hyperglycemia is generally an incidental finding on routine laboratory or point-of-care testing.

Moderate hypoglycemia ([BG] 41-70mg/dL, [2.3-3.9 mmol/L]) is also clinically silent. In conscious patients, severe hypoglycemia ([BG] < 41mg/dL [,2.3 mmol/L]) may result in confusion, agitation, tachycardia, sweating, tremors, seizures and coma. However, in critically ill patients who are unconscious or sedated, even severe hypoglycemia may be clinically silent.

Key management points

• Prevent or treat severe hypoglycemia ([BG] < 41 mg/dL; <2.3 mmol/L).

• Prevent or treat moderate hypoglycemia ([BG] 41- 70 mg/dL; 2.3 – 3.9 mmol/L).

• Prevent or treat hyperglycemia ([BG] > 180 mg/dL; >10 mmol/L).

• Minimize [BG] variability.

Severe hypoglycemia ([BG] < 41 mg/dL; <2.3 mmol/L) is a medical emergency and should be treated with either oral or intravenous glucose. Traditionally, 50 ml (25 g) of 50% glucose is administered intravenously. Cease insulin if it is being administered. Recheck [BG] concentration within 20 minutes. If hypoglycemia has occurred following administration of long acting insulin, long acting oral hypoglycemic agents, or in the context of hepatic failure or an insulin secreting tumour, a continuous infusion of concentrated glucose may be required.

Moderate hypoglycemia ([BG] 41-70 mg/dL; 2.3-3.9mmol/L) should be corrected as a matter of urgency by administration of either oral, enteral or intravenous glucose. If in doubt administer intravenous glucose (5-10g as 20% or 50% glucose) and recheck the blood glucose concentration within 20 minutes.

Preventable causes of hypoglycemia include reduction of calorie intake without a reduction in insulin administration and failure to measure blood glucose frequently enough.

Risk factors for severe hypoglycemia include female sex, increased age, lower body mass index, pre-existing diabetes, low admission [BG], increased severity of illness (APACHE II score), severe sepsis, treatment with renal replacement therapy and treatment with insulin targeting normoglycemia (intensive insulin therapy).

Prolonged hyperglycemia during prolonged critical illness is associated with increased morbidity and mortality. [BG] should be monitored regularly (at least every 4 hours; every 2 hours or more frequently if the patient is being treated with insulin) in every critically ill patient. Unless the patient is eating, in which case hyperglycemia should be investigated and managed as for a patient with diabetes, hyperglycemia associated with critical illness should be managed with intravenous short acting insulin.

The optimum target range for [BG] is not established but current recommendation is to keep [BG] less than 180mg/dL (<10.0 mmol/L), and when insulin is administered to achieve this, [BG] should be maintained between 144-180mg/dL (8.0-10.0 mmol/L). When insulin is being administered [BG] should be measured at least every 2 hours, more frequently if [BG] is decreasing rapidly or if [BG] is close to hypoglycemic range.

The best way to maintain [BG] within a chosen target range is not established but as a general rule protocols with computerized decision support systems (CDSS) perform better than paper based protocols.

2. Emergency Management

Emeregency management of severe hypoglycemia ([BG] <41 mg/dL; <2.3 mmol/L)

• Take confirmatory blood sample and send to laboratory but initiate emergency treatment without waiting for result.

• Discontinue insulin infusion if running.

• If patient is able to drink administer glucose containing drink and check for rapid resolution of symptoms and signs, repeat [BG] measurement after 15-20 minutes, give further glucose containing drink or intravenous glucose if required.

• If patient is unable to drink administer 20-50 mL of 50% glucose and recheck [BG] after 15-20 minutes.

• Ensure patient has ongoing source of calories, enteral nutrition where possible, otherwise consider either parenteral nutrition or intravenous glucose.

• Recheck [BG] every 30 minutes until stabilized above 70 mg/dL (>3.9 mmol/L).

Check list for management of severe hypoglycemia

• Was hypoglycemia confirmed by a second blood sample taken prior to treatment?

• Was cause of hypoglycemia established? Common causes listed below:

  [BG] not measured frequently enough – should be measured at least every 2 hours in patient treated with an intravenous insulin infusion.

  Calorie source discontinued but insulin infusion continued at same rate (insulin should be discontinued or infusion rate reduced whenever patient made NPO/NBM for procedure or extubation).

  Staff error in insulin administration.

  Liver failure or as part of terminal event in dying patient.

  Artefact – bedside point-of-care meters may give falsely low (or falsely high) readings.

  Rewarming from therapeutic hypothermia.

• Is patient receiving an ongoing source of calories in form of enteral nutrition, intravenous glucose or parenteral nutrition?

• Has resolution of hypoglycemia been confirmed?

• Is patient at risk of recurrent hypoglycemia? At risk patients include:

  Patients who have received long acting insulin or oral hypoglycemic agent.

  Patients who have liver failure.

  Patients who are not receiving an ongoing source of calories.

• Has blood glucose been checked every 30 minutes until stabilized above 70mg/dL (3.9 mmol/L)?

3. Diagnosis

Diagnostic criteria and tests

Hypoglycemia and hyperglycemia should be diagnosed by reliable measurement of blood glucose concentration, ideally using either an arterial blood gas (ABG) machine or central laboratory assays.

Hospital laboratory and blood gas machine measurements of blood glucose measurements should be accurate to within 5 mg/dL (0.3 mmol/L) or less. The accuracy of bedside point of care “glucometers” is much more variable, for many the 95% limits of agreement span more than 36 mg/dL (2.0 mmol/L).

Glucose concentration is lower in red blood cells than in plasma; as a result [BG] measured in plasma is approximately 10% higher than that measured in lysed whole blood. The accuracy of some glucometers is affected by hematocrit, arterial oxygen tension and the presence of interfering substances. Some newer glucometers measure and correct for hematocrit and interfering substances but some point-of-care devices cannot distiguish between maltose and glucose and will return falsely high [BG] readings in patients receiving maltose, which is contained in some immunoglobulin preparations and is a product of the metabolism of some peritoneal dialysis fluids.

In ICUs where glucometers are used, it is essential that those using glucometers and interpreting the results know the characteristics of the particular glucometer being used.

Normal lab values

Normal blood glucose concentration: approximately 82-110 mg/dL (4.4-6.1 mmol/L).

Hypoglycemia: less than 71 mg/dL (<3.9 mmol/L).

Severe hypoglycemia: less than 41 mg/dL (<2.3 mmol/L).

(Figure 1)

Figure 1.

4. Specific Treatment

Management of severe hypoglycemia ([BG] < 41 mg/dL; <2.3mmol/L)

Treat this as a medical emergency:

• Take confirmatory blood sample and send to laboratory but initiate emergency treatment without waiting for result.

• Discontinue insulin infusion if running.

• If patient is able to drink administer glucose containing drink and check for rapid resolution of symptoms and signs, repeat [BG] measurement after 15-20 minutes, give further glucose containing drink or intravenous glucose if required.

• If patient unable to drink administer 20-50 mL of 50% glucose and recheck [BG] after 15-20 minutes.

• Ensure patient has ongoing source of calories, enteral nutrition where possible; otherwise consider either parenteral nutrition or intravenous glucose.

• Recheck [BG] every 30 minutes until stabilized above 70 mg/dL.

Management of Moderate hypoglycemia ([BG] 41-70mg/dL; 2.3 – 3.9 mmol/L)

Treat this as a matter of urgency:

• Administer either oral, enteral or intravenous glucose. If in doubt administer intravenous glucose (5g-10g as 20% or 50% glucose) and recheck the blood glucose concentration within 20 minutes.

• Consider reducing or stopping insulin infusion if running.

• Repeat [BG] measurement after 30 minutes, give further glucose containing drink or intravenous glucose if required.

• Ensure patient has ongoing source of calories, enteral nutrition where possible; otherwise consider either parenteral nutrition or intravenous glucose.

• Recheck [BG] every 30 minutes until stabilized above 70 mg/dL (above 3.9mmol/L).

Management of blood glucose in the absence of hypoglycemia

The overall goal of management is to prevent or treat hyperglycemia while minimizing blood glucose variability and without causing hypoglycemia. The optimum blood glucose concentration for critically ill patients has not been established. Current recommendations suggest keeping it below 180 mg/dL (10.0 mmol/L) and when insulin is needed to do so, to target a [BG] of 144-180 mg/dL (8.0-10.0 mmol/L).

Factors affecting blood glucose concentration include:

• Pre-existing diabetes.

• Severity of illness and individual patient’s degree of insulin resistance.

• Administration of glucose enterally or parenterally, separately or as part of a feeding regimen.

• Administration of insulin or oral hypoglycemic agents.

• Administration of drugs increasing insulin resistance (for example, corticosteroids, epinephrine [adrenaline] and other ß2 adrenoceptor agonists, glucagon).

• Change in metabolic rate (e.g. when rewarming from therapeutic or spontaneous hypothermia).

Comprehensive approach to blood glucose

Although administration of insulin is usually central to control of blood glucose in critically ill patients, other considerations are:

• Avoid repeated changes in rate of carbohydrate administration:

  use a feeding protocol that minimizes stopping and starting of enteral feeding.

  avoid administration of drugs using glucose containing fluids.

• consider administering corticosteroids by continuous infusion.

Approaches to insulin administration

Route of administration:

• Intermittent subcutaneous injection (short, intermediate and long acting insulins).

• Intermittent intravenous injection (short acting insulins).

• Continuous intravenous infusion (short acting insulins).

Most research in critically ill patients has been of continuous intravenous administration of short acting insulin. This is appropriate for patients who are receiving a continuous source of calories either enterally or parenterally or for patients who are not being fed but who are insulin resistant due to their illness or pre-existing diabetes. In patients who are eating or drinking, blood glucose may fluctuate dramatically with meals or other oral intake. In such patients hyperglycemia that merits treatment should be treated with either oral hypoglycemic agents or intermittent subcutaneous insulin.

Protocolized insulin administraion

Approaches to the administration of insulin vary from complete nursing discretion or simple paper-based protocols to highly complex computerized algorithms. Most research has focused on protocols to deliver intensive glucose control and in that context computerized protocols have performed better than paper based protocols.

• There have been few, if any, head to head comparisons of individual protocols.

• There is no agreed standard by which to assess protocols. Reported measures of glycemic control inlcude the following:

  time to get [BG] into the target range.

  percentage of time [BG] is in target range.

  mean [BG] during time of study.

  time adjusted measures of degree of hyperglycemia (hyperglycemic index – the area under the glucose curve above 108 mg/dl (6.0 mmol/l) divided by the length of ICU stay).

  indices of degree by which each [BG] measurement differs from target (glycemic penalty index).

  measures of [BG] variability (standard deviation, mean amplitude of glycemic excursion, daily minimum to maximum excursion).

  incidence of severe hypoglycemia.

• Assessment of protocols has generally been limited to effects on blood glucose concentration, not effects on patients’ outcomes.

• Different protocols have not been compared.

• No one protocol has been proven to result in improved patient outcomes.

Refractory cases

Insulin resistance varies markedly between patients and in some patients [BG] may not be controlled despite very large doses of insulin. There are no reliable data to guide treatment in such patients, options are to accept a higher [BG], continue to increase insulin dose or to reduce carbohydrate administration or feeding.

5. Disease monitoring, follow-up and disposition

Response to treatment

The percentage of time patients’ [BG] remains within a set target range depends on the target range chosen. When targeting 80-110mg/dL (which is not currently recommended), most protocols result in patients spending less than 50% of time with [BG] within the target range. This percentage increases if a wider or higher range is targeted.

Prognosis

As patients recover from critcal illness, [BG] tends to return to normal for the individual patient. Patients who have more pronounced hyperglycemia when critically ill have a significantly increased risk of subsequently developing type 2 diabetes. Whether critical illness unmasks patients already at risk of diabetes or hyperglycemia occurring due to critical illness causes or contributes to later development of diabetes is currently unknown.

Most research and controversy surrounds the effect of glucose control on survival of critically ill patients. Most trials have compared intensive glucose control (target [BG] <110mg/dL [6.1 mmol/L]) with conventional control (target [BG] <180 or <216 mg/dL [<10 or <12 mmol/L]). To date more than 30 trials have recruited in excess of 14,000 patients. The trial data can be summarized as follows:

• Overall, there is no increase in survival when intensive glucose control is pursued.

• A beneficial effect of intensive control was reported in Van den Berghe’s original study in patients treated in a surgical ICU; her subsequent study in pediatric patients and in three small trials conducted in China.

• Van den Berghe’s study in medical ICU patients reported reduced duration of mechanical ventilation, ICU and hospital stay but no survival benefit (the reported improved survival in patients staying more than 3 days in the ICU is offset by increased mortality in the remaining patients and Van den Berghe and colleagues could not identify the long stay patients in advance).

• Three mutlicenter trials conducted in Germany, Europe and Australia, New Zealand and North America have collectively suggested that intensive glucose control increases mortality.

• All trials have reported that patients treated with intensive control are at increased risk of severe hypoglycemia ([BG} <40 mg/dL [<2.2mmol/L]) – overall the risk is increased approximately six-fold.

Pathophysiology

Diabetes of injury

Acute illness or injury results in insulin resistance, glucose intolerance and thus hyperglycemia. Hyperglycemia is the end product of a number of processes. Acute illness increases glucose production by the liver with continued gluconeogenesis despite increased endogenous insulin release. Uptake of glucose by skeletal and cardiac muscle, which is insulin dependent, is impaired, whereas insulin independent uptake in other tissues is increased.

A number of counterregulatory hormones and inflammatory cytokines are responsible for “diabetes of injury”; these include catecholamines, cortisol, glucagon, growth hormone and interleukins. Other factors contributing to hyperglycemia during acute illness include immobility, administration of glucose containing fluids and parenteral nutrition, and exogenous administration of catecholamine and corticosteroids.

Causes of hyperglycemia in the critically ill

• Insulin resistance in the liver and peripheral tissues.

• Increased glucose production in the liver and kidney.

• Large glucose loads from feeds and intravenous fluids.

• Administration of drugs increasing insulin resistance.

Putative mechanisms of glucose toxicity

• Intracellular acidosis and reduced ATP level.

• Reduced polymorphonuclear leukocyte chemotaxis.

• Impaired T-cell mediated immune responses.

• Reduced IL-1 and NF-kB expression.

• Decreased cerebral blood flow.

• Pro-inflammatory and pro-oxidant effects.

• Increased IL-6 and IL-10 production.

• Increased endothelial activation.

• Increased expression of tissue factor, leading to a pro-inflammatory and pro-coagulant state.

Adverse effects of hypoglycemia

Although it is well recognized that prolonged severe hypoglycemia can cause permanent neurological injury and death, the adverse effects of transient and more moderate hypoglycemia are less well recognized. They include:

• Impaired autonomic function.

• Alteration of blood rheology.

• White blood cell activation.

• Vasoconstriction.

• Release of inflammatory mediators and cytokines

• Increase in QTc interval.

Although not proven, these mechanisms might explain the strong independent link between moderate and severe hypoglycemia and death.

Adverse effects of increased glucose variability

Increased glucose variability is associated with increased mortality and although a causal relationship is not established, it has been linked to potentially harmful physiological changes such as increased oxidative stress, oxidative DNA damage and pancreatic beta cell dysfunction.

Epidemiology

Hyperglycemia

Hyperglycemia occurs in almost all patients treated in intensive or critical care units. The exact percentage affected depends on the threshold chosen to define hyperglycemia; if a threshold [BG] of 110 mg/dL (6.1 mmol/L) is chosen then around 97% of patients will suffer hyperglycemia. In the NICE SUGAR study, 69% of patients randomized to conventional control were treated with insulin as their [BG] exceed 180mg/dL (10.0 mmol/L). In Van den Berghe studies, where patients randomly assigned to conventional treatment received insulin when their [BG] exceeded 215 mg/dL (12.0 mmol/L), 39% of patients treated in the surgical ICU and 70% of patients treated in the medical ICU received insulin.

Hypoglycemia

The incidence of hypoglycemia is dependent on the threshold used to define the condition and the blood glucose target in the population. The American Diabetes Association defines hypoglycemia as a blood glucose concentration of 70 mg/dL (<3.9 mmol/L) or less whereas most of the focus in critically ill patients has been on severe hypoglycemia, defined as [BG] of 40 mg/dL (2.2 mmol/L) or less.

In the NICE SUGAR study, moderate hypoglycemia ([BG] 41-70 mg/dL; 2.3-3.8 mmol/L) occurred in 75% of patients assigned a [BG] target of 81-108 mg/dL (4.5 – 6.0 mmol/L) and 16% of patients assigned a target of 180 mg/dL (10.0 mmol/L) or less; severe hypoglycemia ([BG] 40 mg/dL; 2.2 mmol/L) occurred in 6.9% and 0.5% of patients, respectively. Across a number of clinical trials the risk of severe hypoglycemia in patients assigned to intensive glucose control averaged around 11% compared with 1.6% for patients assigned to conventional glucose control.

Risk factors for severe hypoglycemia

• Treatment with intensive insulin therapy.

• Female sex.

• Increased age.

• Lower body mass index.

• Pre-existing diabetes and prior treatment with insulin.

• Low [BG] on ICU admission.

• Increased severity of illness (e.g. APACHE II score).

• Severe sepsis.

• Treatment with renal replacement therapy.

Prognosis

As patients recover from critical illness, [BG] tends to return to normal for the individual patient. Patients who have more pronounced hyperglycemia when critically ill have a significantly increased risk of subsequently developing type 2 diabetes. In a study of critically ill medical patients where 140mg/dL (7.8mmol/L) was used as the threshold to define hyperglycemia, 3.5% of those who were not hyperglycemic developed type 2 diabetes within 5 years, compared to 17.1% of those who were hyperglycemic.

Whether critical illness unmasks patients already at risk of diabetes or hyperglycemia occurring due to critical illness causes, or contributes to later development of diabetes, is currently unknown. The impact of different levels of [BG] control during critical illness on later development of diabetes is also currently unknown, although long term follow up of patients randomized to different intensities of glucose control could provide an answer.

Special considerations for nursing and allied health professionals.

In most health care systems critical care nurses are responsible for monitoring blood glucose and treating both hypoglycemia and hyperglycemia. Critical care nurses should expect treating physicians to give clear and unambiguous instructions regarding the blood glucose target for each patient, the frequency of blood glucose measurements and instructions to administer, adjust or cease treatment with insulin dependent on the blood glucose concentration and currently administered therapy.

These instructions may be in the form of a paper-based or computerized protocol. Nursing practice can substantially affect the risk of both hypoglycemia and hyperglycemia. Therefore, the following preventable causes of hypoglycemia should be considered:

• Blood glucose concentration not measured frequently enough.

• Calorie source reduced or discontinued without reduction or cessation of insulin administration (e.g. feeding stopped for a procedure, intrahospital transport or in expectation patient will be extubated).

• Reduction or cessation of parenteral nutrition (including during change of bags).

• Errors in blood glucose measurement (e.g. blood samples drawn from glucose contaminated lines or poor technique when using point-of-care devices).

• Administration of substances that interfere with glucose estimation (e.g. some glucometers cannot differentiate between maltose and glucose, and some intravenous immunoglobulin preparations contain maltose).

Preventable causes of blood glucose fluctuations

• Intravenous administration of medication in glucose containing solutions.

• Rapid intravenous administration of corticosteroids.

What's the evidence?

Hyperglycemia is associated with increased mortality and morbidity in critically ill patients

Krinsley, JS. “Association between hyperglycemia and increased hospital mortality in a heterogeneous population of critically ill patients”. Mayo Clin Proc. vol. 78. 2003. pp. 1471-78. (Retrospective review of 1,826 patients treated in single ICU in USA. Average blood glucose concentration during ICU stay was independently associated with mortality.)

Capes, SE, Hunt, D, Malmberg, K, Gerstein, HC. “Stress hyperglycaemia and increased risk of death after myocardial infarction in patients with and without diabetes: a systematic overview”. Lancet. vol. 355. 2000. pp. 773-8. (Systematic review of 15 cohort studies. Following myocardial infarction, stress hyperglycemia was associated with an increased risk of in-hospital mortality; the risk of congestive heart failure or cardiogenic shock was increased in patients without diabetes.)

Capes, SE, Hunt, D, Malmberg, K, Pathak, P, Gerstein, HC. “Stress Hyperglycemia and Prognosis of Stroke in Nondiabetic and Diabetic Patients: A Systematic Overview”. Stroke. vol. 32. 2001. pp. 2426-32. (Meta-analysis of 26 cohort studies. Acute hyperglycemia predicted increased in-hospital mortality in nondiabetic patients with ischemic stroke and poor functional recovery in nondiabetic stroke survivors.)

Hypoglycemia – risk factors and association with increased mortality in critically ill patients

Vriesendorp, TM, van Santen, S, Devries, JH. “Predisposing factors for hypoglycemia in the intensive care unit”. Crit Care Med. vol. 34. 2006. pp. 96-101. (Retrospective cohort study of 2,272 patients treated in single ICU in The Netherlands. Risk factors for hypoglycemia defined as a blood glucose concentration <45 mg/dL (<2.5 mmol/L) were use of bicarbonate-based substitution fluid during continuous venovenous hemofiltration, a decrease of nutrition without adjustment of insulin infusion, a prior diagnosis of diabetes mellitus, sepsis, and need for inotropic support.)

Hermanides, J, Bosman, RJ, Vriesendorp, TM. “Hypoglycemia is associated with intensive care unit mortality”. Crit Care Med. vol. 38. 2010. pp. 1430-1434. (Retrospective review of 5,961 patients treated in single ICU in The Netherlands. Hypoglycemia defined as a blood glucose concentration, <45 mg/dL (<2.5 mmol/L). After correcting for severity of disease using the daily Sequential Organ Failure Assessment score, age, sex, cardiothoracic surgery, sepsis and diabetes mellitus, the occurrence of hypoglycemia as significantly associated with risk of death.)

Egi, M, Bellomo, R, Stachowski, E. “Hypoglycemia and outcome in critically ill patients”. Mayo Clin Proc. vol. 85. 2010. pp. 217-24.. (Retrospective review of 4,946 patients admitted to two ICUs in Australia. Mild or moderate hypoglycemia (blood glucose below 81 mg/dL; 4.5 mmol/L) was independently associated with mortality. More severe hypoglycemia was associated with greater risk of death.)

Arabi, YM, Tamim, HM, Rishu, AH. “Hypoglycemia with intensive insulin therapy in critically in patients: predisposing factors and association with mortality”. Crit Care Med. vol. 37. 2009 Sep. pp. 2536-44. (Cohort study within 523 patient randomized controlled trial conducted in single hospital in Saudi Arabia. Risk factors for hypoglycemia (defined as blood glucose <40 mg/dL or <2.2 mmol/L) were treatment with intensive insulin therapy, female sex, diabetes, severity of illness, mechanical ventilation, continuous renal replacement therapy and longer ICU length of stay. That hypoglycemia was not an independent risk factor for death may be due to the small size of the study.)

Krinsley, JS, Grover, A. “Severe hypoglycemia in critically ill patients: risk factors and outcomes”. Crit Care Med. vol. 35. 2007. pp. 2262-7. (Retrospective case control study of 102 patients with severe hypoglycemia (blood glucose < 40 mg/dL, <2.2 mmol/L) treated in single ICU in USA. Risk factors for severe hypoglycemia were diabetes, septic shock, renal insufficiency, mechanical ventilation, severity of illness and treatment with tight glycemic control. Severe hypoglycemia was independently associated with increased risk of death.)

Relationship between blood glucose variability and outcome

Egi, M, Bellomo, R, Stachowski, E, French, CJ, Hart, G. “Variability of Blood Glucose Concentration and Short-term Mortality in Critically Ill Patients”. Anesthesiology. vol. 105. 2006. pp. 244-52. (Retrospective review of database of 168,337 blood glucose measurements performed on 7,049 patients in ICU in single hospital in Australia. Both mean and standard deviation (SD) of blood glucose were independently associated with death in the ICU and death in hospital.)

Hermanides, J, Vriesendorp, TM, Bosman, RJ, Zandstra, DF, Hoekstra, JB. “Glucose variability is associated with intensive care unit mortality”. Crit Care Med.. vol. 38. 2010. pp. 838-42. (Prospective cohort study of 5,728 patients treated in ICU in single hospital in The Netherlands. Assessed the relationship of mean absolute glucose change per hour and SD of glucose concentration with ICU and in-hospital death; increases in both measures were associated with increased risk of death.)

Meyfroidt, G, Keenan, DM, Wang, X, Wouters, PJ, Veldhuis, JD. “Dynamic characteristics of blood glucose time series during the course of critical illness: effects of intensive insulin therapy and relative association with mortality”. Crit Care Med.. vol. 38. 2010. pp. 1021-9. (Retrospective analysis of data from two randomized controlled trials conducted in single hospital in Belgium. In multivariable logistic regression analysis with adjustment for baseline risk factors, three measures of blood glucose variability (increased mean daily blood glucose excursion, higher standard deviation of blood glucose and higher jack-knifed approximate entropy) were all independently associated with hospital mortality.)

Measurement methods for blood glucose concentration

Hoedemaekers, CW, Klein Gunnewiek, JM, Prinsen, MA, Willems, JL, Van der Hoeven, JG. “Accuracy of bedside glucose measurement from three glucometers in critically ill patients”. Crit Care Med.. vol. 36. 2008. pp. 3062-6. (Compared the accuracy of blood glucose measurements taken on three glucometers with glucose oxidase-based laboratory measurement; the 95% confidence interval for the mean difference was between 17 and 29 mg/dL [0.94 and 1.60 mmol/L].)

Finkielman, JD, Oyen, LJ, Afessa, B. “Agreement Between Bedside Blood and Plasma Glucose Measurement in the ICU Setting*”. Chest. vol. 127. 2005. pp. 1749-1751. (Retrospective study of agreement between bedside glucometers measurements and hospital laboratory measurement performed at same time. The mean difference between the two techniques was reported as 7.9 mg/dL (0.44 mmol/L), and the limits of agreement were +43.1 to -27.2 mg/dL [+2.39 to -1.51 mmol/L].)

Scott, MG, Bruns, DE, Boyd, JC, Sacks, DB. “Tight Glucose Control in the Intensive Care Unit: Are Glucose Meters up to the Task?”. Clin Chem. vol. 55. 2009. pp. 18-20. (Commentary on accuracy of point-of-care glucose meters.)

Management of blood glucose; impact of intensive insulin therapy / intensive glucose control on outcome

Van den Berghe, G, Wouters, P, Weekers, F. “Intensive insulin therapy in critically ill patients”. N Engl J Med. vol. 345. 2001. pp. 1359-67. (First trial to randomly assign critically ill patients to either normoglycemia or a higher range of [BG]. Conducted in the surgical ICU of a quaternary referral hospital in Belgium. 62% of the patients were admitted to the SICU after cardiac surgery. In-hospital mortality was 7.2% for patients assigned to normoglycemia, versus 10.9% for patients assigned a [BG] target of less than 215mg/dL (12.0 mmol/L). Hypoglycemia ([BG] less than 40 mg/dL (<2.2mmol/L) occurred in 5.2% of those assigned normoglycemia compared to 0.8% of the conventional care group.)

Van den Berghe, G, Wilmer, A, Hermans, G. “Intensive Insulin Therapy in the Medical ICU”. N Engl J Med. vol. 354. 2006. pp. 449-61. (Second trial by Van den Berghe et al using same methodology but conducted in their medical ICU. Randomly assigned 1,200 patients expected to stay in the MICU for 3 or more days to normoglycemia or higher [BG] target. In-hospital mortality was not significantly improved by targeting normoglycemia. Patients assigned normoglycemia had reduced morbidity. Hypoglycemia ([BG] <40 mg/dL [<2.2mmol/L]) occurred in 18.7% of those assigned normoglycemia compared to 3.1% of the conventional care group (data in electronic supplement).

Vlasselaers, D, Milants, I, Desmet, L. “Intensive insulin therapy for patients in paediatric intensive care: a prospective, randomised controlled study”. The Lancet. vol. 373. 2009. pp. 547-56. (Third trial by Van den Berghe’s group which randomly assigned critically ill infants and children aged under 16 years to age-adjusted normoglycemia or a [BG}] of less than 216 mg/dL (12.0 mmol/L). 75% of the infants or children were admitted to the ICU following surgery for congenital heart disease. Duration of stay in the ICU, the primary outcome measure, was reduced by targeting normoglycemia. Mortality at 30 days was reduced from 5.1% to 2.3%. Hypoglycemia ([BG] less than 40 mg/dL (<2.2mmol/L) occurred in 24.9% of those assigned normoglycemia compared to 1.4% of the conventional care group.

“Intensive versus conventional glucose control in critically ill patients”. N Engl J Med. vol. 360. 2009. pp. 1346-9. (International multi-center trial that randomly assigned 6,104 critically ill adults admitted to an ICU in one of 42 hospitals in Australia, New Zealand, Canada or the USA to normoglycemia ([BG] target 81-108 mg/dL [4.5-6.0 mmol/L]) or a target of 180 mg/dL (10.0 mmol/L) or less. 37% of the patients were admitted to the ICU after surgery, the majority of those (64.2%) after emergency surgery. In patients assigned to normoglycemia, mortality was significantly increased (27.5% vs. 24.9%) at 90 days. Hypoglycemia ([BG] less than 40 mg/dL (<2.2mmol/L) occurred in 6.8% of those assigned normoglycemia compared to 0.5% of the conventional care group.)

Wiener, RS, Wiener, DC, Larson, RJ. “Benefits and risks of tight glucose control in critically ill adults: a meta-analysis”. JAMA. vol. 300. 2008. pp. 933-944. (Meta-analysis of 29 randomized controlled trials totaling 8,432 patients conducted before the results of the NICE SUGAR trial were known. Hospital mortality did not differ between tight glucose control and usual care overall. Mortality not significantly different when trials were stratified according to target glucose concentration in tight control arm (very tight <110mg/dL [<6,1 mmol/L] or moderately tight <150 mg/dL [<8.3 mmol/L]) or when stratified by type of ICU (surgical versus medical versus mixed). Tight glucose control was associated with decreased risk of septicemia but five fold increase in risk of hypoglycemia.)

Griesdale, DE, de Souza, RJ, van Dam, RM. “Intensive insulin therapy and mortality among critically ill patients: a meta-analysis including NICE-SUGAR study data”. CMAJ. vol. 180. 2009. pp. 821-7. (Meta-analysis of 26 trials involving a total of 13,567 patients including the data from the NICE SUGAR study. Overall reported no benefit from intensive glucose control. There was a six-fold increase in the risk of severe hypoglycemia with intensive glucose control.)

Marik, PE, Preiser, JC. “Toward understanding tight glycemic control in the ICU: a systematic review and metaanalysis”. Chest. vol. 137. pp. 544-51. (Meta-analysis of seven trials involving a total of 11,425 patients. Tight glucose control did not reduce mortality, bloodstream infections or use of renal replacement therapy. Incidence of hypoglycemia significantly increased by tight glucose control. Meta-repression suggested a relationship between the effect of tight glucose control and number of calories delivered parenterally. This effect was due to the results of the trials conducted in Leuven where most calories were delivered parenterally.)

Protocolized insulin administration:

Eslami, S, Abu-Hanna, A, de Jonge, E, de Keizer, N. “Tight glycemic control and computerized decision-support systems: a systematic review”. Intensive Care Med. vol. 35. 2009. pp. 1505-17. (Systematic review of computerized decision support systems designed to target normoglycemia.)

Hoekstra, M, Vogelzang, M, Verbitskiy, E, Nijsten, MW. “Health technology assessment review: Computerized glucose regulation in the intensive care unit – how to create artificial control”. Crit Care. vol. 13. 2009. pp. 223(Review of computerized decision support systems designed to target normoglycemia.)

Blaha, J, Kopecky, P, Matias, M, Hovorka, R. “Comparison of three protocols for tight glycemic control in cardiac surgery patients”. Diabetes Care.. vol. 32. 2009. pp. 757-61. (Randomized assessment of three protocols in cardiac surgical patients.)

Measuring glycemic control

Chase, JG, Shaw, GM. “Is there more to glycemic control than meets the eye?”. Crit Care. vol. 11. 2007. pp. 160(Commentary summarizes issues relating to evaluating protocols and measuring glycemic control.)

Pathophysiology

Dungan, KM, Braithwaite, SS, Preiser, JC. “Stress hyperglycaemia”. Lancet.. vol. 373. 2009. pp. 1798-807. (Review article providing classification of stress hyperglycemia and discussion of mechanisms by which hyperglycemia may cause harm.)

Brealey, D, Singer, M. “Hyperglycemia in critical illness: a review”. J Diabetes Sci Technol.. vol. 3. 2009. pp. 1250-60. (Review of pathophysiology of stress-induced hyperglycemia.)

Prognosis

Gornik, I, Vujaklija-Brajkovic, A, Renar, IP, Gasparovic, V. “A prospective observational study of the relationship of critical illness associated hyperglycaemia in medical ICU patients and subsequent development of type 2 diabetes”. Crit Care.. vol. 14. 2010. pp. R130(Describes prospective followup of ICU patients without history of impaired glucose tolerance for a minimum of 5 years. Dichotomized patients according to [BG] in ICU. Those with [BG] in ICU of 140 mg/dL (7.8 mmol/L) or more were more than five times more likely to develop diabetes during followup.)

Nursing considerations

Aragon, D. “Evaluation of nursing work and perceptions about blood glucose testing in tight glycemic control”. Am J Crit Care.. vol. 15. 2006. pp. 370-7. (Estimated nursing workload of targeting normoglycemia at 2 hours per patient per day.)

Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.

No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.