1. Define diabetes mellitus and review management of acute glycemic complications.

2. Cite the available literature addressing inpatient glycemic control and patient outcomes.

3. Discuss intraoperative glucose management.

4. Summarize the risks and benefits of blood transfusion and examine transfusion thresholds.

5. Discuss intraoperative management that would reduce exposure to allogeneic blood and review transfusion-related acute lung injury.

Perioperative Glycemic Management
Daniel R. Brown, MD, PhD, Assistant Professor of Anesthesiology, and Chair, Division of Critical Care Medicine, Mayo Clinic, College of Medicine, Rochester, MN

Diabetes mellitus: abnormal carbohydrate metabolism resulting in hyperglycemia; most common endocrinopathy

Classification of diabetes

Type 1: destruction of pancreatic β cells; primarily autoimmune, but in idiopathic version, β cells destroyed in absence of circulating antibodies; typically occurs in younger patient; however, variants of adult-onset type 1 diabetes can occur; absolute insulin deficiency

Type 2: most common; variable degrees of insulin deficiency and resistance; subtypes include drug-induced and gestational

Diagnosis

Current American Diabetes Association definitions: fasting glucose ≥126 mg/dL or random glucose >200 mg/dL with symptoms (eg, polydipsia, polyuria); glucose level >200 mg/dL after 2-hr oral glucose challenge

“Impaired” glycemic control: fasting glucose 100 to 125 mg/dL; possibly reversible with appropriate intervention

Acute glycemic complications: hypoglycemia—have mechanism in place for immediate treatment; hyperglycemia— diabetic ketoacidosis (DKA); nonketotic hyperosmolar states (NKHS)

Diabetic ketoacidosis: earlier mortality rates ≥80%, but current mortality 5% to 10%; hyperglycemia, dehydration, hyperosmolarity, and increased anion gap common; ketones associated with Kussmaul breathing, fruity breath, nausea, and vomiting; hyperkalemia frequently present, as well as sodium, magnesium, and phosphorus deficits; potential spurious hyponatremia (for every 100 mg/dL that glucose level above 100, add 1.5 to 2.0 mEq to measured sodium level); management—routine monitors plus arterial and central venous access; intravenous (IV) insulin and crystalloid; start dextrose infusion when glucose ≈200 mg/dL; continue IV insulin until ketones cleared and acidemia resolved; supplement electrolytes

Nonketotic hyperosmolar states: typically seen in type 2 diabetes; patient typically more dehydrated, more hyperglycemic, and more hyperosmolar than DKA patient; neurologic manifestations common (eg, confusion, coma, focal neurologic deficits); patients lack ketone production but may have lactic acidosis; potential for thrombotic events; treatment—insulin therapy; volume resuscitation with crystalloids; make sure correction gradual, due to concerns about cerebral edema; frequent neurologic evaluations

Hyperglycemia

Adverse outcomes: hyperglycemia independent risk factor; increases rate of hospitalization and length-of-stay (LOS); also increases infectious complications; Latham (2001) found diabetes and postoperative hyperglycemia associated with development of surgical-site infections

Inpatient mortality: Umpierrez (2002) found newly discovered hyperglycemia associated with highest mortality

Stroke prognosis: Capes (2001) performed systematic review of 26 studies; reported admission glucose >144 mg/dL associated with increased risk for in-hospital and 30-day mortality; in survivors, hyperglycemia associated with poor functional recovery; Malmberg (1999) studied 620 diabetic patients with acute myocardial infarction (MI); patients randomly assigned to insulin drip or routine antidiabetic therapy; intensive insulin therapy showed significantly improved outcomes

Insulin therapy: Furnary (2003) and colleagues looked at glucose managed with subcutaneous insulin and insulin infusion; reported reduced absolute and risk-adjusted mortality of 57% and 50%, respectively

Benefits of intensive insulin therapy in critically ill patients: van den Berghe (2001) compared strict glycemic control (80-110 mg/dL) to conventional therapy (treat for glucose >215 mg/dL, maintain at 180-200 mg/dL); reduced intensive care unit (ICU) mortality from 8.0% to 4.6%; in-hospital mortality reduced by one-third; reduced mortality attributable to effect on patients in ICU >5 days; bloodstream infections reduced by almost 50%; similar study repeated in 1200 medical patients found no difference in mortality by intention-to-treat analysis (40.0% conventional vs 37.7% intensive); decreased morbidity in intensive group (LOS, weaning from mechanical ventilation, and renal failure); increased mortality if ICU LOS <3 days, decreased mortality if ICU LOS ≥3 days

Intraoperative blood glucose control: Ouattara (2005) reported poor intraoperative glycemic control associated with severe morbidity; Gandhi (2005) performed retrospective study of >400 patients undergoing cardiac surgery; primary end point composite of death and major morbidity; primary end point more common in men, older patients, diabetics, and those receiving insulin during surgery; mean and maximal intraoperative glucose significantly associated with poor outcomes; other studies report similar findings linking hyperglycemia and death during cardiopulmonary bypass (CPB), before carotid endarterectomy, during first 48 hr after infrainguinal bypass, and during noncardiac nonvascular surgery; Gandhi (2007) found intensive intraoperative glycemic control does not reduce perioperative death or morbidity; increased incidence of death and stroke raises concern about routine implementation of this intervention; Butterworth (2005) looked at neurologic outcomes after cardiac surgery; hyperglycemic control during CPB did not improve neurologic or neurobehavioral outcomes in patients with diabetes

Intensive insulin therapy and inflammation: analysis of C-reactive protein (CRP) concentrations in van den Berghe trials; baseline CRP elevated but decreased with time in ICU; decrease more pronounced in intensive insulin therapy; glycemic control and other factors (eg, ventilator management) have impact on inflammatory response

Metrics: controlled postoperative blood glucose values on postoperative days 1 and 2 drawn closest to 6:00 AM

Current Benefits and Risks of Blood Transfusion
Michael A. Gropper, MD, PhD, Professor and Vice Chair, Department of Anesthesia and Perioperative Care; Director, Critical Care Medicine, and Chair for Medical Quality, University of California, San Francisco, School of Medicine

Blood component therapy: whole blood donated then spun off into various components; freezing and thawing results in factor VIII-poor plasma and cryoprecipitate (used in specific situations for specific factor deficiencies)

Packed red blood cells (RBCs): hematocrit ≈70%; fresh-frozen plasma (FFP) likely sold separately; tends to be anticoagulated with citrate

Platelet concentrates: stored at room temperature for ≤4 days; most commonly contaminated blood product

FFP: used commonly for coagulopathy; titration better than with other blood products; prothrombin time (PT), partial thromboplastin time (PTT), and international normalized ratio (INR) useful in deciding whether to administer; use only when necessary

Cryoprecipitate: almost always unnecessary; may be useful in hemophilia or von Willebrand’s factor deficiency; usually type-specific

Transfusion thresholds: point at which risk associated with not transfusing becomes greater than risk associated with transfusing; controversy exists over optimum hemoglobin (Hb); problems include immediate and delayed transfusion reactions, infection (eg, viral infections), and immunosuppression; however, patient with coronary disease or carotid disease may be at risk for stroke or MI with no transfusion

Complications of transfusion (noninfectious): febrile nonhemolytic transfusion reaction; anaphylaxis (rare); graft vs host disease; fluid overload; transfusion-related acute lung injury (TRALI; leading cause of death from transfusion in United States)

Immune suppression by transfusion: modifications in T-cell and B-cell function; down-regulation of antigen-presenting cells; studies suggest transfusion recipients tend to have higher risk for postoperative infection and recurrence of tumor (particularly colorectal cancer)

Transfusion and perioperative infection: Koval study (1997) of hip arthroplasty found rate of postoperative infection 27% in transfused patients vs 15% in patients not transfused; Houbiers study (1997) of colorectal surgery found those requiring transfusion had higher infection rate (39% vs 24%); 60% increase in risk for 1 to 3 U of blood

Universal leukoreduction: Hébert (2003) tracked mortality, pneumonia, bacteremia, and other complications that might be attributed to transfusion before and after universal leukodepletion (Canada, August 2000); results seemed to favor leukoreduction; however, may be other changes in practice that occurred over same period that led to reduction; ultimately resulted in 10% to 20% fewer complications once universal leukodepletion completed; based on these findings, Food and Drug Administration (FDA) and American Association of Blood Banks (AABB) agreed to similar measures in United States

Transfusion triggers: Adams and Lundy (1942) indicated transfusion necessary when Hb concentration reaches 8 to 10 g/dL; transfusion trigger is Hb value below which risk associated with not transfusing (decreased O2 carrying capacity) exceeds risk associated with transfusion; physiologically tolerable Hb not equal to optimal Hb (varies with coexisting disease)

Transfusion practice guidelines: American College of Physicians—no automatic threshold value; transfuse on unit-by-unit basis; National Institutes of Health (NIH)—Hb trigger of 7 g/dL; American Society of Anesthesiologists—transfusion rarely indicated if Hb >10 g/dL, and almost always indicated if Hb <6 g/dL; College of American Pathologists—almost always transfuse for Hb <6 g/dL or for excessive bleeding, but, ideally, use invasive monitoring to decide

Isovolemic anemia: Weiskopf (1998) studied 11 healthy patients and 21 volunteers; Hb decreased to 5 g/dL; measured O2 consumption, plasma lactate, and ST changes; also measured increase in heart rate, stroke volume, and cardiac index; found that patients “actually did quite well” (no lactic acidosis; no significant electrocardiography [ECG] changes); however, Hogue (1998) found anemia with tachycardia does cause problems; Carson (1996) studied patients who were Jehovah’s Witnesses who refused transfusion yet underwent surgery; if no coronary disease and preoperative Hb of 6 g/dL, risk for death doubled; however, with coronary disease, risk for death increased 16-fold; Wu (2001) studied ≈80 000 patients with MI (≥65 yr of age); looked at impact of transfusion and anemia; 90% survived ≤30 days with hematocrit (Hct) 39% to 48%; if Hct 5% to 24%, only 65% survived for 30 days (but transfusion resulted in 80% improvement in survival); Hb of 10 g/dL and Hct of 30% appear to be threshold for benefit

Intraoperative cell salvage: reinfusion of blood lost from surgical field after washing; good efficacy demonstrated in orthopedic surgery, borderline in cardiac surgery; cost-effectiveness depends on blood loss

Hemostatic drugs: aminocaproic acid (eg, Amicar) and tranexamic acid known as lysine analogues; block binding of plasminogen to fibrin; reduce fibrinolysis and stabilize clot; used routinely in “redo” cardiac surgery; efficacious in low doses; increase risk for thrombosis; aprotinin previously widely used in cardiac surgery but now controversial; appears to be increasing mortality; factor VIIa popular, but most hospitals have restricted usage due to expense; increases thrombin expression at site of bleeding; large meta-analysis found large increase in thrombotic events

Transfusion-related acute lung injury: incidence—1 in 5000 blood components; 1 in 400 platelet U; 25 million U of blood products given annually; 5000 cases of TRALI per year; 1000 attributable deaths per year (leading cause of transfusion-related mortality); additional comments—first described in 1985; found antibodies in ≈90% of 36 cases; tended to be anti-HLA antibodies from implicated donor; definitions—acute respiratory distress syndrome (ARDS) indicates oxygen defect, PaO2 /FiO2 <200 mm Hg on 100% O2 , bilateral infiltrates, and no evidence of congestive heart failure; acute lung injury (ALI) same as above except PaO2 /FiO2 <300 mm Hg; consensus definitions—patients without ALI risk factors have onset of signs or symptoms within 6 hr of transfusion; develop lung injury; patient with signs or symptoms of ALI who receives transfusion and then gets much worse more difficult to identify (usually seen in operating room [OR]); clinical course—majority of cases present within 1 to 2 hr of transfusion; most recover within 72 hr, but ≈10% die, and 10% to 15% have prolonged course in ICU; diagnosis—clinical; diagnosis of exclusion; difficult to differentiate transfusion-associated circulatory overload (TACO) from TRALI; likely that patients have components of both; clinical presentation—dyspnea; pulmonary edema; hypoxemia; tachycardia; fever; hypotension or hypertension; decreased white blood cell (WBC) count (occurs in ≈60% of cases); TACO—occurs in ≈10% of transfusions; diastolic dysfunction and pulmonary edema result; difficult to differentiate from TRALI; differential diagnosis can include use of pulmonary artery (PA) catheter, B-type natriuretic peptide (BNP), or echocardiography; causes—inflammatory response in lungs “stirs up” WBC in pulmonary circulation; macrophages activated, release cytokines that break down surfactant, and cause pulmonary edema and protein exudation into lung; antibody hypothesis—donor blood contains antibody to recipient WBC antigens; may also be caused by recipient antibodies to donor WBC antigens; antibodies may be HLA or against other WBC antigens; antibodies cause “mayhem in the lung” leading to pulmonary edema; most cases associated with antibodies to WBC antigens; seems to occur in OR and ICU; neutrophil priming hypothesis—neutrophils primed by infection or surgery, then activated by transfusion; giving older blood products may be one of major triggers of TRALI; many cases of TRALI without antibodies; female donor screening—study in abdominal aortic aneurysm (AAA) repair found male donor FFP reduces risk for TRALI by 60%; in United States, females eliminated from donating plasma or platelets; treatment—low tidal volume mechanical ventilation; steroids not shown to be efficacious; prevention—universal leukocyte reduction; female plasma donor exclusion; wash cellular components (particularly older RBCs) using cell salvage; use of fresher blood and platelets; use of recombinant products