Acute Kidney Injury


Acute Kidney Injury or AKI is the new term for acute renal failure. This new terminology emphasizes the fact that kidney injury is a broad spectrum of functional derangement, and is significant no matter how minor the decrease in function. The term “renal failure” is often associated with the most severe degrees of impairment and can encourage a sense of complacency when dealing with “lesser” degrees of dysfunction. The simple fact is that any degree of renal impairment is associated with poorer patient outcome and cannot be ignored.

AKI is common in hospitalized patients, and considered especially so in the critically ill. Traditional incidence figures are problematic because of wide variations in the definitions used across studies (see AKIN and RIFLE scores in the section on Severity Scoring below), and this has led to the development of research definitions that are clinically applicable and that take into account all grades of AKI, not just the severest forms and the use of renal replacement therapies. In a study of almost 30,000 ICU patients, the prevalence of AKI (using strict criteria of UOP < 200 mL in 12 hours and/or BUN > 84 mg/dL) was 5.7%. This equates to AKIN stage 3 AKI or RIFLE F. The most common contributing factor to ARF was septic shock (47.5%). Only 30% of patients had documented preadmission renal dysfunction [JAMA 294: 813, 2005].

Types of AORF

Prerenal Failure

Responsible for 30-40% of cases in the ICU, but probably more than 40% in the operating room. Caused by decreased effective arterial volume which can be due to hypovolemia, cardiac dysfunction, or loss of vascular tone, but also due to renal vasoconstriction (ex. NSAIDS) or ↓ glomerular filtration pressure (ex. ACE-I/ARB) as well as renal artery stenosis. FENA < 1% and BUN/Cr > 20

Intrinsic Renal Failure

ATN alone causes 50% of AORF in the ICU, probably significantly less in the operating room. In the ICU, intrinsic renal failure is primarily due to ATN (prolonged prerenal failure, sepsis, shock, toxin), AIN, and acute glomerulonephritis.


ATN is most common, most often secondary to sepsis, circulatory shock, and/or nephrotoxins such as dye, aminoglycosides, or myoglobin [Crit Care Med 24: 192, 1996]. Often ATN is associated with multi-organ injury and thus the treatment regimen chosen should take this into account. The mortality rates from ATN in hospitalized and ICU patients are about 37.1% and 78.6%, respectively [Chest 128: 2847, 2005]. Look for FENA > 2% (kidney can no longer retain sodium) as well as muddy casts, +/- RBCs in urine.


AIN can be difficult to distinguish from ATN, thus always look at the medication list. Almost any drug can cause AIN, and unfortunately fever, rash, and/or eosinophilia may not be present. Worse, renal injury can occur months or even years after therapy [New Horiz 3: 608, 1995]. Oral prednisone at 0.5 – 1.0 mg/kg daily for 1-4 weeks may speed recovery [Crit Care Clin 22: 357, 2006] Myoglobinuric renal failure is usually mild. Diagnose with a positive urine dipstick in the absence of erythrocytes on microscopy. Rhabdomyolysis should be suspected if azotemia increases faster than possible by renal shutdown alone (ie BUN > 30 mg/dL, Cr > 2 mg/dL, K+ > 0.5 mEq/L, or HCO3- > 2 mEq/L). This can be confirmed by measuring serum CPK, LDH, or aldolase (specific for skeletal muscle)

Drugs That Commonly Cause AIN


Small Vessel Disease

Can lead to ARF, especially emboli, HUS/TTP/DIC, preeclampsia, and even malignant hyperthermia


Rare to manifest acutely in OR/ICU population, will see dysmorphic RBC and RBC casts on urinalysis

Postrenal Failure

Any obstruction to kidney outflow. Rare in the OR/ICU patient unless only one kidney is present


Risk Factors after non-cardiac surgery

What are the risk factors for kidney injury after noncardiac surgery? If we know these factors, can the risk be modifiable? It is already known that anemia and transfusions are associated with kidney injury after cardiac surgery. Is the same true for patients who do not undergo cardiac surgery? Dr. Michael Walsh, Departments of Medicine and Clinical Epidemiology and Biostatistics, McMaster University, E Hamilton, ON and co-authors studied whether preoperative hemoglobin and perioperative reduction of hemoglobin were associated with kidney injury in patients who underwent noncardiac surgery. The results of their study were published in Anesthesia & Analgesia in the article “The Association Between Perioperative Hemoglobin and Acute Kidney Injury in Patients Having Noncardiac Surgery” [Walsh M et al.].


Diagnosis (1. microscopy 2. [Na+]urine 3. FENa) Urine microscopy is mandatory but only helpful in identifying intrinsic renal causes – look for epithelial cells and casts (ATN), white cell casts (nephritis), pigmented casts (myoglobinuria). If this is undiagnostic, move on to [Na+]urine. Low sodium is relatively specific for prerenal failure, but not sensitive – if [Na+]urine < 20 mEq/L, this is most likely prerenal, however [Na+]urine as high as 40 mEq/L have been found in prerenal patients (ie those on diuretic therapy or with an ongoing intrinsic process as well). FENa is probably the best way to distinguish the two, however it is underutilized because of its cumbersome nature (FENa = {[Na]urine/[Na]plasma}/{[Cr]urine/[Cr]plasma}). If FENa < 1.0% the problem is most likely prerenal, but if FENa > 2.0% it is probably intrinsic, although occasionally there are exceptions. Note that a normal FENa is < 1.0%. Also, a 2-fold increase in Cr indicates renal injury, and a 3-fold increase indicates renal failure. According to Stein, Urine Na >30mEq/L suggests intrinsic renal pathology (ex. acute tubular necrosis), urine Na <10mEq/L suggests volume depletion, and anything in between requires a FENa.

Maximum Rate of Change 2° to Renal Shutdown in 24 hours

BUN: 20 – 30 mEq/dL
Creatinine: 1 – 2 mg/dL
[K+]: 0.3 – 0.5 mEq/L
[HCO3-]: 1 – 2 mEq/L

Note: if these values are exceeded, suspect rhabdomyolysis

Electrolyte Abnormalities


K+ outflow is responsible for phase III (repolarization) of the cardiac cycle. Insulin and beta agonists promote the influx of K+. Total body stores include 4200 mEq intracellularly, versus only 12 mEq in the plasma. Potassium excretion of K+ is determined primarily by [K+] and aldosterone levels.

Hyperkalemia in Renal Failure: Anesthetic Concerns

Time Course of Hyperkalemia

Potassium balance usually maintained in early renal failure, thus hyperkalemia is a late sign (a GFR of 8 cc/min is adequate to clear potassium).

Hyperkalemia and Induction

In non-renal failure patients, induction with SCh increased [K+] by 0.4 mEq/L [Manninen PH et al. Anesth Analg 70: 172, 1990] and by 0.09 mEq/L [Stacey MR et al. Anaesthesia 50: 933, 1995].  ; Anesth Analg 95: 119, 2002]. Schow et al. reviewed 40,000 anesthetics at Duke University Medical Center, and found 38 cases in which SCh was used with a starting potassium > 5.5 mEq/L. No fatalities or arrhythmias occurred, and the authors estimated that the maximal risk of an event was 7.9% or less [Schow et al.].

Duration of Neuromuscular Blockade

Despite the fact that muscle weakness is a sign of hyperkalemia (at > 7 mEq/L), Miller at al. showed that the prolongation of neuromuscular blockade following paralysis in renal failure patients is not related to potassium levels, but rather to the relative inability of the kidneys to excrete NMBDs or their metabolites.

Signs of Hyperkalemia

(> 6 mEq/L): peaked T waves, prolonged PR interval (ex. AV block), flattened P wave, QRS widened to sine wave; (> 7 mEq/L): musculoskeletal weakness; Note that these effects are worsened by hyponatremia, hypocalcemia, and acidosis, all of which are present in the setting of renal failure.

Renal Failure: Hyperkalemia

  • Causes: decreased excretion (lowered GRF, ACE-inhibitors), intracellular release (metabolic acidosis), increased intake (transfusion)
  • Signs: peaked T waves, prolonged PR interval, flattened P wave, QRS widened to sine wave, musculoskeletal weakness
  • Acute Treatment: 1g CaCl2, 50g glucose, 10 U insulin, furosemide, 100 mEq bicarbonate (controversial)
  • Anesthesia Concerns: very real but may be overstated; increase in non-renal failure patients < 0.5 mEq/L with SCh. Affect on paralysis not related to K+

Other Electrolytes

Renal failure is often complicated by elevations in potassium, phosphate, and magnesium and decreases in sodium and calcium. Additionally, chronic renal failure patients often present with an anion gap metabolic acidosis. Urea, creatinine, uric acid, sulfate, phosphate, phosphorus, lipids, cholesterol, neutral fats, and some amino/organic acids may accumulate, while albumin levels fall.

Renal Failure: Electrolytes

  • Elevated Electrolytes: potassium, phosphate, and magnesium
  • Decreased Electrolytes: sodium, calcium
  • Other Increases: urea, creatinine, uric acid, sulfate, phosphate, phosphorus, lipids, cholesterol, neutral fats, and some amino/organic acids
  • Other Decreases: albumin
  • Acid/Base: anion gap metabolic acidosis


Advances over the last 50 years have been sparse at best. The avoidance of further nephrotoxic insults, such as nonsteroidal antiinflammatory drugs, nephrotoxic antibiotics, and radiocontrast media is very important, as are correction reversible causes. Maintenance of hemodynamics and renal perfusion are cornerstones of care [Ann Intern Med 379: 744, 2002] Filling pressures should be normal and not have decreased by 4 mm Hg (CVP at least 4 mm Hg, wedge 8 mm Hg). If not, provide volume. Cardiac output should be checked next – if it’s low, give more volume until CVP > 10 mm Hg and wedge approaches 20 mm Hg. If, after full volume resuscitation, cardiac output is still inadequate, proceed with ionotropic intervention (start dobutamine 5 μg/kg/min) to obtain a CI of at least 3L/min/m2. If volume and cardiac output are corrected and renal failure persists, it is likely intrinsic and cannot be effectively treated Despite Marino’s recommendations, a prospective randomized controlled trial is needed to clarify whether the augmentation of perfusion pressure can achieve improvement in renal function or even prevent the occurrence of renal dysfunction [Chest 128: 2847, 2005]. Studies have also shown that increasing cardiac output and oxygen delivery through the administration of large volumes of fluid and inotropic agents, and aggressive RBC transfusion may increase mortality [NEJM 330:1717, 1994; NEJM 333: 1025, 1995; Crit Care Med 24:517, 1996; NEJM 340: 409, 1999] and thus the perfusion benefits must be weighed against the risks.

Low dose dopamine should NOT be used [New Horiz 3: 650, 1005; Heart Lung 22: 171, 1993; Am J Med 1: 49, 1996] as while it does improve renal blood flow and promote diuresis, it has never shown to improve renal function or outcomes and may cause bowel ischemia [Lancet 344: 7, 1994]. Bellomo and colleagues reported on 328 critically ill patients with ARF who were randomly assigned to continuous infusion of placebo or low-dose dopamine (2 ug/kg/min). Peak serum creatinine concentration, requirement for dialysis, length of hospital stay, and mortality rate did not differ between groups [Lancet 356: 2139, 2000]

Data on furosemide are equivocal – less than 10% of diuretics actually reach the tubules in renal failure [Clin Pharmacol Ther 40: 134, 1986] – if diuresis is attempted, it should be continuous IV infusion and not a bolus, as infusion has been shown to be more effective [Crit Care Med 22: 1323, 1994]. A prospective, randomized, placebo-controlled, double-blind study examining the effect of loop diuretics on renal recovery, dialysis, and death in patients with ARF found no effect [Nephrol Dial Transplant 12: 2592, 1997]. More recently, a prospective, multicenter epidemiologic study by Uchino examined the impact of diuretics on critically ill patients with ARF and found that their use was not associated with higher mortality[Crit Care Med 32: 1669, 2004]. Therefore, it is reasonable to administer a single trial of furosemide in escalating doses, and if the patients does not respond, the drug should not be readministered.

Fenoldopam has showed promise in small studies but is as of yet unproven – one study showed that when fenoldopam was used prophylactically in patients undergoing aortic surgery, its use was associated with improvement in renal function, and reductions in dialysis requirements, length of hospital stay, and mortality [J Cardiovasc Pharmacol Ther 6: 31, 2001]

Volume resuscitation with monitoring of [K+] and [PO42-] levels is essential. Alkalinization is rarely necessary. 30% of these patients will require hemodialysis [Am J Surg 188: 801, 2004] Hemofiltration devices are improving rapidly, and while their clinical utility is not yet known, they may soon be a viable option in critical patients [Crit Care 9: 523, 2005] – neither CAVH or CVVH cause hypotension, although CAVH cannot be used in hypotensive patients as it relies on an arterial pressure gradient.

Prophylaxis Before Contrast

During the past 5 years, 19 randomized controlled trials, 4 prospective nonrandomized studies, and 11 meta-analyses that explored the role of acetylcysteine for prevention of contrast-induced nephropathy have been published. A recent meta-analysis demonstrated that these 34 empirical studies have not yet conclusively resolved this question [Arch Intern Med 166: 161, 2006]. Another meta-analysis of randomized data concerning NAC before coronary angiography to prevent CIN in patients with impaired renal function was neither conclusive nor provided proof beyond a reasonable doubt to influence clinical practice and public policy [Am Heart J 151: 140, 2006]

Hydration is the primary intervention for preventing contrast nephropathy [Nephrol Dial Transplant 14: 1064, 1999]. Different regimens of saline hydration have been used, but no one regimen has demonstrated clear superiority. Bader randomized patients undergoing CT or digital angiography to receive either 2,000 mL of IV fluid over 24 hours (12 hours before and 12 hours after contrast) or 300 mL of IV fluid during the radiologic procedure. GFR rate fell by 18.3 mL/minute and 34.6 mL/minute respectively (p < .05), suggesting that slow hydration is superior to bolus expansion during the procedure [Clin Nephrol 62: 1, 2004]. Still, as of yet no sufficiently powered, controlled, prospective trials have examined the minimally effective length of time, optimal rate, and fluid composition of intravenous hydration required before and after contrast administration in high-risk azotemic patients [Clev Clin J Med 73:75, 2006]

A prospective, single-center, RCT of 119 patients with stable serum creatinine levels of at least 1.1 mg/dL received a 154-mEq/L infusion of either sodium chloride (n=59) or sodium bicarbonate (n=60) before and after iopamidol administration (370 mg iodine/mL). Patients received 154 mEq/L of either sodium chloride or sodium bicarbonate, as a bolus of 3 mL/kg per hour for 1 hour before iopamidol contrast, followed by an infusion of 1 mL/kg per hour for 6 hours after the procedure. The primary end point of contrast-induced nephropathy occurred in 8 patients (13.6%) infused with sodium chloride but in only 1 (1.7%) of those receiving sodium bicarbonate (mean difference, 11.9%; 95% confidence interval [CI], 2.6%-21.2%; P=.02). A follow-up registry of 191 consecutive patients receiving prophylactic sodium bicarbonate and meeting the same inclusion criteria as the study resulted in 3 cases of contrast-induced nephropathy (1.6%; 95% CI, 0%-3.4%) [JAMA 291: 2328, 2004]. In a similar study of 111 patients undergoing emergency PCI – 56 patients received an infusion of sodium bicarbonate plus N-acetylcysteine (N-AC) started just before contrast injection and continued for 12 h after PCI. 55 other patients received the standard hydration protocol consisting of intravenous isotonic saline for 12 h after PCI. In both groups, 2 doses of oral N-AC were administered the next day. A creatinine concentration >0.5 mg/dl from baseline was observed in 1 patient given bicarbonate/NAC pre-procedurally, and in 12 patients who only received hydration (AORF was observed in 1 patient vs. 7 patients, p = 0.032) [J Am Coll Cardiol. 49: 1283, 2007]

In a recent meta-analysis, the incidence of contrast-induced nephropathy ranged from 2% to 26% in patients receiving N-acetylcysteine plus sodium chloride and 11% to 45% in those patients administered sodium chloride hydration alone [Lancet 362: 598, 2003]. There are some who do not believe that NAC is truly helpful. Based on two articles [Crit Care Clin 21: 193, 2005; J Gen Intern Med 20: 193, 2005], Marino recommends giving 100-150 cc/hr fluids 3-12 hours before the procedure (or 300-500 cc for emergent procedures), as well as 600 mg NAC bid starting 24 hours before and ending 24 hours after the procedure.