Acute Kidney Injury in Icu

Jonathan E. Sevransky, MD, MHS, Section EditorConcise Denitive Review

Acute kidney injury in the intensive care unit: An update andprimer for the intensivist

Paula Dennen, MD; Ivor S. Douglas, MD; Robert Anderson, MD

Acute kidney injury (AKI), pre-viously termed acute renal fail-ure, refers to a sudden declinein kidney function causing dis-turbances in uid, electrolyte, and acidbase balance because of a loss in smallsolute clearance and decreased glomerularltration rate (GFR). The nomenclatureshift to AKI more accurately represents thespectrum of disease from subclinical injury tocomplete organ failure. This review focuseson key questions for the intensivist faced withAKI in the intensive care unit (ICU).

Epidemiology of AKI in the ICU

AKI in the ICU is common, increasingin incidence (14), and is associated witha substantial increase in morbidity and

mortality (5, 6). AKI occurs in approxi-mately 7% of all hospitalized patients (7)and in up to 36% to 67% of critically illpatients depending on the denition used(6, 811). Based on 75,000 critically illadults, more severe AKI occurs in 4% to25% of all ICU admissions (6, 8, 9, 11).On average, 5% to 6% of ICU patientswith AKI require renal replacement ther-apy (RRT) (6, 811).

Reported mortality in ICU patientswith AKI varies considerably betweenstudies depending on AKI denitionand the patient population studied(e.g., sepsis, trauma, cardiothoracicsurgery, or contrast nephropathy). Inthe majority of studies, mortality in-creases proportionately with increasingseverity of AKI (6, 1013). In patientswith severe AKI requiring RRT, mortal-ity is approximately 50% to 70% (9,1416). While AKI requiring RRT in theICU is a well-recognized independentrisk factor for in-hospital mortality(17), even small changes in serum cre-atinine (SCr) are associated with in-creased mortality (18 21). Notably,multiple studies of patients with AKIand sepsis (2224), mechanical ventila-tion (25), major trauma (26, 27), car-diopulmonary bypass (17, 2830), andburn injuries (31) have consistentlydemonstrated an increased risk of death

despite adjustment for comorbiditiesand severity of illness.

Morbidity, a less appreciated conse-quence of AKI in the ICU, is associatedwith increased cost (18), increased lengthof stay (6, 14, 18, 26), and increased riskof chronic kidney disease (CKD), includ-ing end-stage kidney disease (9, 15, 16,3237). The true incidence of CKD afterAKI is unknown because epidemiologicstudies do not routinely or consistentlyreport rates of renal recovery and thosethat do use variable denitions (38).

Denition of AKI in the ICU

More than 35 denitions of AKI cur-rently exist in the literature (39). TheAcute Dialysis Quality Initiative convenedin 2002 and proposed the RIFLE classi-cation (risk, injury, failure, loss, end-stage kidney disease) specically for AKIin critically ill patients (Table 1) (40).Using SCr and urine output, the RIFLEcriteria dene three grades of severityand two outcome classes. The most se-vere classication met by either criterionshould be used. Of note, patients withprimary kidney diseases such as glomer-ulonephritis were excluded from this def-inition.

More recently the Acute Kidney InjuryNetwork (AKIN), an international multi-disciplinary organization composed of

From Divisions of Nephrology and Critical CareMedicine (PD), Division of Pulmonary Sciences andCritical Care Medicine (ISD), and Department of Med-icine (RA), Denver Health Medical Center and Univer-sity of Colorado, Denver, CO.

Denver Health Medical Center and University ofColorado, Denver, CO, are Acute Respiratory DistressSyndreome network investigation sites (PD and ISD).

The authors have not disclosed any potential con-icts of interest.

For information regarding this article, E-mail:[email protected]

Copyright 2009 by the Society of Critical CareMedicine and Lippincott Williams & Wilkins

DOI: 10.1097/CCM.0b013e3181bfb0b5

Objective: Acute kidney injury is common in critically illpatients and is associated with signicant morbidity and mor-tality. Patients across the spectrum of critical illness haveacute kidney injury. This requires clinicians from across dis-ciplines to be familiar with recent advances in denitions,diagnosis, prevention, and management of acute kidney injuryin the intensive care unit. The purpose of this concise review,therefore, is to address, for the non-nephrologist, clinicallyrelevant topical questions regarding acute kidney injury in theintensive care unit.

Data Sources: The authors (nephrologists and intensivists)performed a directed review of PubMed to evaluate topicsincluding the denition, diagnosis, prevention, and treatmentof acute kidney injury in the intensive care unit. The goal of

this review is to address topics important to the practicingintensivist.

Data Synthesis and Findings: Whenever available, preferentialconsideration was given to randomized controlled trials. In theabsence of randomized trials, observational and retrospectivestudies and consensus opinions were included.

Conclusions: Acute kidney injury in the intensive care unit is aclinically relevant problem requiring awareness and expertiseamong physicians from a wide variety of elds. Although manyquestions remain controversial and without denitive answers, aperiodic update of this rapidly evolving eld provides a frameworkfor understanding and managing acute kidney injury in the inten-sive care unit. (Crit Care Med 2010; 38:261275)

KEY WORDS: acute kidney injury; intensive care unit

261Crit Care Med 2010 Vol. 38, No. 1

nephrologists and intensivists, furthermodied the RIFLE criteria recognizingthat even very small changes in SCr(0.3 mg/dL) adversely impact clinicaloutcome (6, 7, 10, 11, 19, 21, 41). Accord-ing to AKIN, the most current consensusdiagnostic criteria for AKI is an abrupt(within 48 hrs) reduction in kidney func-tion currently dened as an absolute in-crease in serum creatinine of more thanor equal to 0.3 mg/dL (26.4 mol/L), apercentage increase in serum creatinineof 50% (1.5-fold from baseline), or areduction in urine output (documentedoliguria of 0.5 mL/kg/hr for 6 hrs)(42). Importantly, the AKIN denitionand classication system incorporatescreatinine, urine output, and time (Table1). Both the RIFLE and AKIN criteriawere developed to facilitate clinical inves-tigation and comparison across studypopulations. Epidemiologic data compar-ing the RIFLE and AKIN criteria havedemonstrated concordance in critically illpatients (43, 44).

Diagnosis of AKI in the ICU

Traditional tools to diagnose AKI(SCr) and determine etiology of AKI(clinical history, physical examination,renal ultrasound, fractional excretion ofsodium [FeNa], fractional excretion ofurea, blood urea nitrogen [BUN], andurine microscopy) remain the corner-stone of diagnostic tools available to theclinician in the ICU. The use of SCr toestimate GFR is limited, however, by thelack of steady-state conditions in criti-cally ill patients. Determinants of the SCr(rate of production, apparent volume of

distribution, and rate of elimination) arevariable in the ICU setting (6, 811, 45,46). Medications (e.g., trimethoprim,cimetidine) impair creatinine secretionand therefore may cause increases in SCrwithout reecting a true decrease inGFR. Finally, SCr lacks sensitivity andunderestimates the degree of kidney dys-function in a critically ill patient. In-creases in SCr substantially lag behind areduction in GFR (Fig. 1) and thus do notprovide a useful real-time assessment ofGFR.

AKI spans the continuum from prere-nal azotemia to acute tubular necrosis,from functional to structural injury. Ef-forts to differentiate between these twoentities have classically included FeNaand urine microscopy. Urine microscopycan be helpful in differential diagnosis(e.g., granular casts and renal tubularepithelial cells in acute tubular necrosis,cellular casts in glomerular injury, eosi-nophiluria in acute interstitial nephritis,or atheroembolic AKI). Of clinical note,nephrologist review of urine microscopy

Figure 1. Relationship between glomerular ltration rate (GFR) and serum creatinine (SCr). Largechanges in GFR (e.g., 50% decrease from 120 mL/min to 60 mL/min) are reected in only smallchanges in SCr (0.7 mg/dL to 1.2 mg/dL).

Table 1. Classication/staging systems for acute kidney injury

RIFLE SCr Criteria UOP CriteriaAKINStage SCr Criteria UOP Criteria

R 1 SCr 1.5 0.5 mL/kg/hr 6 hrs 1 1 in SCr 0.3 mg/dL or 1150% to 200% frombaseline (1.5- to 2-fold)

0.5 mL/kg/hr for 8 hrs

I 1 SCr 2 0.5 mL/kg/hr 12 hrs 2 1 in SCr to 200% to 300%from baseline(2- to 3-fold)

0.5 mL/kg/hr for 12 hrs

F 1 SCr 3, or SCr 4 mg/dLwith an acute rise of at least0.5 mg/dL

0.5 mL/kg/hr 24 hrsor anuria 12 hrs

3 1 in SCr to 300% (3-fold)from baseline or SCr 4mg/dL with an acute rise ofat least 0.5 mg/dL

0.5 mL/kg/hr 24 hrs oranuria 12 hrs

L Persistent loss of kidney functionfor 4 wks

E Persistent loss of kidney functionfor 3 months

RIFLE, risk, injury, failure, loss, end-stage kidney disease; AKIN, acute kidney injury network; SCr, serum creatinine; UOP, urine output.RIFLE criteria adapted from Bellomo et al (40). AKIN criteria adapted from Mehta et al (42).

262 Crit Care Med 2010 Vol. 38, No. 1

has been demonstrated to be superior toclinical laboratory interpretation (47).Using a proposed scoring system, micro-scopic examination of the urine sedimentis a highly predictive method for differ-entiating prerenal azotemia from acutetubular necrosis (48). However, the pres-ence of muddy brown casts and renaltubular epithelial cells are usually seenrelatively late and thus are not sensitivefor early detection of AKI (49, 50). FeNa isfrequently useful for differentiating pre-renal (diminished renal perfusion, FeNa1%) from intra-renal (ischemia ornephrotoxins, FeNa 2%) (50, 51). Urinemicroscopy and FeNa can be valuabletools in determining the cause of AKI buthave no current role in early detection ordiagnosis of AKI. Furthermore, prere-nal and intra-renal causes of AKI com-monly coexist in the ICU patient.

Prerenal azotemia, in the absence ofvalidated new diagnostic biomarkers, of-ten remains a retrospective diagnosis,made only after response to a volumechallenge. Whereas it is important to ap-propriately identify and treat prerenalazotemia, uid administration is notwithout consequence in the critically illpatient. A complete assessment of the pa-tients overall volume status is pivotalbefore aggressive resuscitative efforts toenhance renal perfusion. This is of par-ticular importance considering data dem-onstrating adverse effects of volume over-load in critically ill patients (52, 53).Because of the limitations of traditionaltools, novel candidate biomarkers of AKI(discussed separately) are being activelyinvestigated.

Common Causes of AKI inthe ICU

The cause of AKI in the ICU is com-monly multi-factorial and frequentlydevelops from a combination of hypovo-lemia, sepsis, medications, and hemody-namic perturbations (Table 2). It is fre-quently not possible to isolate a singlecause, thereby further complicating thesearch for effective interventions in thiscomplex disease process. The pathophys-iology of AKI varies according to the un-derlying etiology and is beyond the scopeof this article.

Sepsis is the most common cause ofAKI in a general ICU, accounting for upto 50% of cases (6, 811, 23, 45, 54). AKIis common after cardiac surgery, occur-ring in up to 42% of patients withoutpre-existing kidney disease, and is associ-

ated with increased morbidity and mor-tality with elevations in SCr as small as0.3 mg/dL (19). Trauma associated AKI ismulti-factorial (e.g., hemorrhagic shock,abdominal compartment syndrome,rhabdomyolysis) and occurs in up to 31%of adult trauma patients (55). The kid-neys are early sensors of intra-abdominalhypertension and abdominal compart-ment pressures 12 mm Hg may be as-sociated with AKI (56). A sustained intra-abdominal pressure 20 mm Hg inassociation with new organ dysfunctionwill be associated with AKI in 30% ofcases (57, 58). Rhabdomyolysis accountsfor 28% of trauma-associated AKI requir-ing dialysis (59).

Medications are a common cause ofAKI and, according to Uchino et al (9),account for nearly 20% of all cases of AKIin the ICU. The mechanism of medicationinduced AKI is variable and includesacute interstitial nephritis, direct tubulartoxicity (e.g., aminoglycosides), and he-modynamic perturbations (e.g., nonste-roidal anti-inammatory agents, angio-tensin-converting enzyme inhibitors).Acute interstitial nephritis is likely anunder-recognized etiology of medication-associated AKI in the ICU because of therelative paucity of clinical ndings andneed for high index of suspicion. Table 3lists common nephrotoxins encounteredin the care of critically ill patients.

Prevention and Management ofAKI in the ICU

Primary prevention of AKI in the ICUis limited to those conditions in whichthe timing of injury is predictable, suchas exposure to radiocontrast dye, cardio-pulmonary bypass, large-volume para-centesis in a cirrhotic patient, or chemo-

therapy. In contrast to most cases ofcommunity-acquired AKI, nearly all casesof ICU-associated AKI result from morethan a single insult (6, 811, 45, 50, 60,61). In the critically ill patient, the rstkidney insult is often not predictable.Therefore, prevention of AKI in the ICUoften means prevention of a secondaryinsult in an at-risk patient. For exam-ple, in a retrospective study of5000 ICUpatients, 67% of patients had AKI de-velop, and 45% of AKI occurred after ICUadmission (6). It is in these patients thatthere is a potential role for prevention.

General principles of secondary AKIprevention include: (1) recognition of un-derlying risk factors that predispose pa-tients to AKI (e.g., diabetes, chronic kid-ney disease, age, hypertension, cardiac orliver dysfunction); and (2) maintenanceof renal perfusion, avoidance of hypergly-cemia, and avoidance of nephrotoxins inthese high-risk patients. Specic clinicalsituations in which there is evidence forpreventive strategies (e.g., contrast expo-sure, hepatorenal syndrome [HRS]) arediscussed.

Preventing Contrast-Induced Ne-phropathy. The primary strategies for con-trast-induced nephropathy (CIN) preven-tion include hydration, N-acetylcysteine(NAC), and use of low-volume nonioniclow-osmolar or iso-osmolar contrast. Nostrategy has been effective in completelypreventing CIN. Risk factors for CIN in-clude diabetes, CKD, hypotension, effec-tive or true volume depletion (includingcirrhosis and congestive heart failure),and concurrent use of nephrotoxic med-

Table 2. Common causes of AKI in the ICU

Five Most Common Causes of AKI in the ICUa

Sepsis (most common) Major surgery Low cardiac output Hypovolemia Medications

Other Common Causes of AKI in the ICU Hepatorenal syndrome Trauma Cardiopulmonary bypass Abdominal compartment syndrome Rhabdomyolysis Obstruction

aThe ve most common causes of acute kid-ney injury (AKI) in the intensive care unit (ICU)based on nearly 30,000 patients (9).

Table 3. Common nephrotoxins that cause acutekidney injury in intensive care unit patients

Exogenous Medications

NSAIDSAntimicrobials

AminoglycosidesAmphotericinPenicillinsa

Acyclovirb

Chemotherapeutic agents Radiocontrast dye Ingestions

Ethylene glycolEndogenous Rhabdomyolysis Hemolysis (HUS/TTP) Tumor lysis syndrome

NSAIDS, non-steroidal anti-inammatorydrugs; HUS, hemolytic uremic syndrome; TTP,thrombotic thrombocytopenic purpura.

aAcute interstitial nephritis (AIN); bcrystalnephropathy.


ications. Critically ill patients intuitivelyrepresent a patient population at highrisk for CIN given frequent hemodynamicinstability, multiple organ dysfunction,use of nephrotoxic medications, and mul-tiple underlying comorbidities (e.g., dia-betes, CKD). However, despite the largenumber of randomized controlled trials(RCT) published on prevention strategiesfor CIN, there has been only one RCTperformed specically in critically illadults (111). The true incidence of andrisk for CIN in critically ill patients isthus unknown.

Adequate volume expression is a well-established measure to decrease the riskof CIN, whereas the choice of uid re-mains controversial. Trials comparingthe use of sodium bicarbonate and so-dium chloride for the prevention of CINhave yielded conicting results. Fivemeta-analyses of sodium bicarbonatesuggest a benecial role of isotonic so-dium bicarbonate over isotonic saline(112116); however, there is considerableheterogeneity and some publication biasconfounding these ndings. The most re-cent RCT of bicarbonate vs. normal salineshowed no difference in the primary out-come of 25% decrement in GFR within4 days (117). Based on currently availableevidence, there is a strong suggestionthat sodium bicarbonate may be superiorto isotonic saline to decrease the risk ofCIN.

NAC is a free radical scavenger shownto decrease the risk of CIN compared toplacebo (118). Since 2003, 10 meta-analyses published on the role of NAC inCIN have yielded conicting results likelyattributable, in part, to heterogeneity inpatient populations. In a recent meta-analysis of 41 studies, NAC plus salinereduced the risk for CIN more effectivelythan saline alone (119). A previous meta-analysis in 2007 by Gonzales et al (120)did not support the efcacy of NAC toprevent or decrease the risk of CIN. Fur-thermore, there are conicting data as towhether NAC, itself, may decrease SCrmeasurement without affecting GFR(121, 122).

Low-volume nonionic low-osmolar oriso-osmolar contrast preparations areclearly associated with a decrease in CINwhen compared to high osmolar agents.The data regarding nonionic low-osmolarcontrast media vs. iso-osmolar contrastmedia (currently only iodixanol) is con-troversial. Two meta-analyses report con-icting results (123, 124). McCullough etal (123) found that use of iso-osmolar con-

trast media resulted in a lower incidence ofCIN when compared to low-osmolar con-trast media. However, Heinrich et al (124),in the most recent meta-analysis, re-ported no signicant difference betweenthe two unless the low-osmolar contrastmedia was iohexol, suggesting that alllow-osmolar contrast media preparationsmay not be the same.

Both small observational and prospec-tive studies have shown an increase in therisk of CIN with peri-procedural use ofangiotensin-converting enzyme inhibi-tors (125127). However, a recent ran-domized prospective trial performed instable outpatients did not show any dif-ference in incidence of CIN between pa-tients who did or did not discontinueangiotensin-converting enzyme inhibi-tors or angiotensin receptor blockers be-fore contrast (128). Angiotensin-convert-ing enzyme inhibitors have not beenprospectively studied in the critically ill.Therefore, although there is currently in-sufcient evidence to support discontin-uation of these medications in criticallyill adults, further study is warrantedgiven the widespread use of these agentsin clinical practice.

Whereas the use of peri-proceduralhemoltration in patients undergoingpercutaneous coronary intervention wasshown, in two studies, to decrease therisk of AKI (5% vs. 50%; p .0001) (129,130), this has not been widely adoptedinto clinical practice. In a systematic re-view of extracorporeal therapies for pre-vention of CIN, analysis of the hemodial-ysis studies alone (including ve RCT),there was no benet of hemodialysis and,in fact, there was a trend favoring stan-dard therapy compared to prophylactichemodialysis (131). A subsequent RCT ofprophylactic hemodialysis in 82 patientswith advanced CKD (baseline SCr 4.9 mg/dL) demonstrated improved outcomes(shorter length of stay and lower rate oflong-term dialysis dependence after hos-pital discharge) with prophylactic hemo-dialysis (132). A critical limitation of allof these studies is that the clinical endpoint SCr was directly impacted by theintervention itself (hemoltration or he-modialysis).

Fenoldopam and theophylline are twoadditional agents that have been consid-ered for their potential role in the pre-vention of CIN. None of the four RCTcomparing fenoldopam to either salinealone (133, 134) or NAC (135, 136) dem-onstrated any benecial effect in the pre-vention of CIN. The role of theophylline

for CIN prevention is inconsistent acrossstudies. Although two meta-analyses sug-gest that prophylactic theophylline mayprovide some benet, the studies wereperformed in primarily low-risk patients,and clinically relevant outcomes were notconsistently reported (137, 138). There-fore, we cannot currently recommend theuse of theophylline for prevention of CINin critically ill patients.

The majority of these studies were notperformed in critically ill patients andtherefore provide no denitive guidanceas to how the risk of CIN in the criticallyill should be ameliorated. Because of theabsence of sufcient data in the patientpopulation of interest, clinicians must ex-trapolate from the best available evidencefrom other patient populations. There-fore, our recommendations include: (1)avoid use of intravenous contrast in high-risk patients if alterative imaging tech-niques are available; (2) use preexposurevolume expansion using either bicarbon-ate or isotonic saline; (3) although ofquestionable benet, use of NAC is safe,inexpensive, and may decrease risk ofAKI; (4) avoid concomitant use of neph-rotoxic medications if possible; and (5)use low-volume low-osmolar or iso-osmolar contrast. Future studies areneeded to determine the true role ofthese preventive measures in critically illpatients.

Preventing AKI in Hepatic Dysfunc-tion. AKI is a common complication ofcritically ill patients with hepatic failure.Pentoxifylline decreases the incidence ofAKI attributable to HRS in acute alco-holic hepatitis (139). Use of intravenousalbumin in patients with cirrhosis andspontaneous bacterial peritonitis signi-cantly reduces both the incidence of AKI(33% to 10%) and mortality (41% to22%) (140). Albumin decreases the inci-dence of AKI after large-volume paracen-tesis (141), and when used in combina-tion with splanchnic vasoconstrictingagents (e.g., terlipressin) may decreasemortality in HRS (142, 143). However,denitive therapy for AKI as a conse-quence of HRS remains liver transplan-tation in appropriate candidates. Fiverandomized trials of vasoconstrictingagents (terlipressin or noradrenalin) plusalbumin in the treatment of HRS all dem-onstrated improved renal function inHRS (144148). A mortality benet wasonly demonstrated in responders to ther-apy (145). Terlipressin is not available inthe US. In a retrospective study per-formed in the US, patients treated with


vasopressin had signicantly higher re-covery rates and improved survival whencompared to octreotide alone (149). Fur-thermore, ndings from three small ob-servational and retrospective studiesdemonstrate improved outcomes withmidodrine and octreotide (HRS reversaland decreased mortality) (150 152).These ndings justify a larger RCT toappropriately evaluate this treatmentmodality.

Management of AKI in the ICU re-volves around optimizing hemodynamicsand renal perfusion, correcting metabolicderangements, providing adequate nutri-tion, and mitigating progression of in-jury. These management considerationsare discussed.

Maintain Renal Perfusion. Optimiza-tion of renal perfusion may require vol-ume resuscitation, inotropic, or vasopres-sor support. Extrapolated primarily fromanimal studies (62, 63), the human kid-ney has a compromised ability to auto-regulate (maintain constancy of renalblood ow and GFR over a wide range ofrenal perfusion pressures) in AKI. There-fore, as a priority, prevention or manage-ment of AKI should include maintenanceof hemodynamic stability and avoidanceof volume depletion. A mean arterialpressure of 65 mm Hg is a generallyaccepted target; however, the data arelimited (64, 65) and do not include pa-tients with established AKI (loss of auto-regulation). The level at which renalblood ow becomes dependent on sys-temic arterial pressure varies signi-cantly based on age, underlying illness(e.g., hypertension), and the acute illnessor condition (AKI, sepsis, and cardiopul-monary bypass). After volume resuscita-tion, blood ow should be restored towithin autoregulatory parameters. Thisfrequently requires vasopressor or inotro-pic support in the setting of septic shock,the most common cause of AKI in theICU. There are currently no RCT compar-ing vasopressor agents; therefore, there isno evidence that, from a renal protectionstandpoint, there is a vasopressor agentof choice to improve kidney outcomes.

Decreased renal blood ow (attribut-able to either hypotension or high renalvascular resistance, from an imbalancebetween renal vasoconstriction and vaso-dilation) is a common feature in manyforms of AKI. Consequently, there hasbeen considerable interest in renal vaso-dilators to maintain renal perfusion forprevention or treatment of AKI. Whereasdopamine infusion may cause a transient

improvement in urine output (66), re-nal dose dopamine does not reduce theincidence of AKI, the need for RRT, orimprove outcomes in AKI (6671). Fur-thermore, low-dose dopamine mayworsen renal perfusion in critically illadults with AKI (72) and is associatedwith increased myocardial oxygen de-mand and an increased incidence of atrialbrillation (73). There is additional con-cern for extrarenal adverse effects of do-pamine, including negative immuno-modulating effects (74). Thus, there isbroad consensus that dopamine is poten-tially harmful and without evidence ofclinical benet for either prevention ortreatment of AKI. Therefore, its contin-ued use for putative renal protectionshould be avoided.

Fenoldopam is a selective dopamine-1receptor agonist approved for the treat-ment of hypertensive crisis (75). Paradox-ically, the lowest doses of fenoldopam(1 g/kg per min) are purported toincrease renal blood ow without sys-temic effects. Despite encouraging datafrom pilot studies, (7678) a prospectiveplacebo-controlled study of low-dosefenoldopam in sepsis failed to decreasemortality or need for RRT despite asmaller increase in SCr (79). Larger stud-ies to validate the meta-analytic observa-tion that fenoldopam both reduces theneed for RRT (OR, 0.54; p .007) anddecreases mortality (OR, 0.64; p .01)(80) are currently ongoing in cardiac sur-gery patients (clinicaltrials.gov ID:NCT00557219).

Fluid Choice in AKI. The primaryphysiologic intention of volume resusci-tation is the restoration of circulatingvolume to prevent or mitigate organ in-jury. The kidneys normally receive up to25% of the cardiac output and are exquis-itely sensitive to hypoperfusion attribut-able to true or relative hypovolemia. Forthis reason, the question of whether aparticular type of uid inuences devel-opment of AKI is of pivotal importance.

Whereas crystalloid solutions remainthe preferred treatment in usual care, thedebate over whether colloid solutionsprovide any additional benet remains anarea of active investigation (8185). In alandmark trial evaluating the impact ofuid choice on clinical outcomes, theSAFE study investigators randomizednearly 7000 patients to volume resuscita-tion with saline or albumin. They dem-onstrated no difference in survival orneed for RRT between the two groups(86). In post hoc subgroup analysis, re-

suscitation with albumin was associatedwith increased mortality in critically illpatients after traumatic brain injury (87).In contrast, there was a trend towardimproved survival in septic shock patientsreceiving albumin (30.7% in albumingroup vs. 35.3% in saline group; p .09)(86). Based on currently available litera-ture, there is no evidence of a mortalitybenet supporting the preferential use ofalbumin over crystalloids in a heteroge-nous critically ill patient population (84).

Synthetic colloids (e.g., hydroxyethylstarches, dextrans) are still widely useddespite multiple reported safety concernswith regard to renal outcomes (8890).An increased risk of AKI with the use ofhydroxyethyl starches has been demon-strated in multiple small studies, andmost recently a systematic review of 12randomized trials demonstrated an in-creased risk of AKI with the use of hy-droxyethyl starches among patients withsepsis (91). In contrast, the largest indi-vidual retrospective analysis (SOAP studycohort, 92) explored the effects of hy-droxyethyl starches on renal function anddid not nd the use of hydroxyethylstarches to be an independent risk factorfor AKI or need for RRT (93). The doseand preparation varied between studies.The adverse event prole has been linked,in part, to the individual preparation,with the lowest molecular weight offeringthe best side effect prole.

The question of uid managementdoes not end with the choice of uid;careful consideration of the amount ofuid administered is also important. Crit-ical illness is a dynamic process requiringfrequent assessment of and adjustment touid status. In a prospective RCT of pa-tients with acute respiratory distress syn-drome, a uid conservative strategy de-creased ventilator days and did notincrease the need for RRT (53). Further-more, an observational study of 3000patients demonstrated an association be-tween positive uid balance and in-creased mortality in patients with AKI(52). However, the question remainswhether this is simply a marker of sever-ity of illness or true causation; this ob-servation warrants further investigation.

Avoid Hyperglycemia. Although thebenecial effects of intensive insulintherapy on mortality in critically ill pa-tients remains controversial (9496), twolarge RCT demonstrated a decreased in-cidence of AKI and a decreased require-ment for RRT with tight glucose control(95, 96). Furthermore, a more detailed


secondary analysis strongly suggests thattight blood glucose control may be reno-protective in critically ill patients (97).Two smaller retrospective studies re-ported similar results (decreased inci-dence of AKI and decreased need for post-operative dialysis) in nondiabetic cardiacsurgical patients (98) and in patients re-ceiving total parenteral nutrition (99).However, in contrast, in the largest andmost recent prospective RCT of intensivevs. conventional glucose control in6000 critically ill patients, there was nodifference in the number of patients re-quiring RRT (94). The overall incidenceof AKI, however, was not reported in thisstudy. It therefore remains unclear ifthere is a reno-protective role for tightglycemic control and, if present, whetherany such effect is attributable to theavoidance of glucose toxicity or a bene-cial effect of insulin. These ndings war-rant further study, especially in view ofthe fact that intensive glycemic controlmay be associated with a higher fre-quency of clinically relevant hypoglyce-mia.

Avoid Nephrotoxins. Nephrotoxicmedications are a contributing factor inup to 25% of all severe AKI in critically illpatients (8, 9; Table 3). Identication ofat-risk patients is pivotal. Aminoglyco-sides, although less commonly used forsevere Gram-negative infections thanpreviously, are associated with signicantnephrotoxicity. Although once-daily dos-ing of aminoglycosides has been shown,in some studies, to decrease the inci-dence of AKI (100, 101), published meta-analyses support comparable efcacy anddecreased cost but do not consistentlydemonstrate a signicant reduction innephrotoxicity (102106). Extended in-terval dosing should not be used in pa-tients with CKD. Standard amphotericinB has been associated with AKI in 25% to30% of patients (107). The lipid formula-tion of amphotericin B is preferred be-cause of reduced nephrotoxicity of 19%vs. 34% (108). Caspofungin, a newer an-tifungal agent, is associated with an evensafer renal prole (109). The use of apro-tinin, a serine protease inhibitor used todecrease blood loss during cardiac sur-gery, has been associated with increasedrisk of AKI and need for dialysis (110).

ICU patients frequently have uctuat-ing renal function and a variable volumeof distribution. Standard estimates of re-nal function are poor in critically ill pa-tients. Therefore, medications must becarefully dose adjusted because of varied

pharmacokinetics in critically ill patientswith and without underlying CKD.

Diuretics in AKI. Use of diuretics inthe prevention or treatment of AKI hasphysiologic merit but its use is not sup-ported by prospective clinical study. Di-uretics can increase urine output buthave not been found to have a consistentimpact on mortality (153157). Mehta etal (157) demonstrated that failure to re-spond to diuretics was associated with anincreased risk of death and non-recoveryof renal function. Subsequently, in alarge, prospective, multinational study,Uchino et al (158) did not demonstrate anincreased mortality, thus leaving unre-solved the therapeutic role of diuretics incritically ill patients with renal dysfunc-tion. Although oliguric AKI has been as-sociated with worse outcomes than nono-liguric AKI (159), there is no evidencesupporting efforts to convert nonoliguricAKI with diuretics. Diuretics have notbeen found to shorten the duration ofAKI, reduce the need for RRT, or improveoverall outcomes (160). Furthermore, arecently published RCT comparing theuse of furosemide vs. placebo in the re-covery phase of AKI requiring continuousrenal replacement therapy (CRRT), furo-semide was found to increase urine out-put and sodium excretion but did notimprove renal recovery (161). In a multi-national survey, nephrologists and inten-sivists reported clinical uncertainty aboutthe use of diuretics in AKI, thus justifyingthe need for a denitive RCT (162).

Because diuretic use in AKI has notbeen shown to decrease mortality, thereis no role for diuretics to convert oliguricAKI to nonoliguric AKI. However, regard-ing an increased appreciation for the po-tential detrimental downstream effects ofvolume overload, it may be reasonable totry diuretics for control of volume over-load. The clinician should, however, becareful not to delay initiation of RRT forvolume overload in the critically ill pa-tient with AKI.

Nutritional Considerations. Malnutri-tion in hospitalized patients is associatedwith increased mortality (163). Assess-ment of the nutritional status of criticallyill patients is limited by the unreliabilityof traditional markers of nutritional sta-tus in critical illness in general, and AKIin particular. Prealbumin is excretedmainly by the kidneys and hence may befalsely elevated in patients with AKI(164). Patients with AKI are hypercata-bolic with a negative nitrogen balance(165), resulting from both increased pro-

tein catabolism and impaired protein syn-thesis.

The impact of CRRT on nutrition inthe ICU is two-fold. Because protein ca-tabolism is markedly increased in mostpatients requiring CRRT (165167), theuse of CRRT enhances the clinicians abil-ity to provide adequate nutrition becauseof an improved ability to manage volume.Unfortunately, the recommendedamount of protein in this population re-mains controversial and recommenda-tions are based solely on expert opinion,because there are no data available fromRCT. Although there are no studies dem-onstrating a benet in outcomes (e.g.,survival or dialysis-free days), consensusrecommendations include nonprotein ca-loric intake of 20 to 30 kcal/kg bodyweight per day and a protein intake of 1.5g/kg per day (168). However, several stud-ies have demonstrated a less negative oreven positive nitrogen balance in thosepatients receiving up to 2.5 g/kg per daywhile receiving CRRT without evidence ofadverse effects (169171). An increase innonprotein calories in critically ill pa-tients with AKI does not improve nitro-gen balance (172).

RRT for AKI in the ICU

Despite decades of clinical trials inves-tigating potential pharmacologic inter-ventions in AKI, current treatment op-tions are primarily limited to RRT.Practice patterns vary widely regardingtiming of initiation of RRT, dose deliv-ered, and choice of modality as evidencedby international surveys (173176).There is no current consensus on theindications for RRT for AKI. With agreater appreciation for and understand-ing of the role of the kidney in distantorgan injury (177), it may be more appro-priate to consider renal replacementtherapy as renal supportive therapy (178).For the purposes of this review, we reviewthe most up-to-date evidence availableaddressing timing, dosing, and modalityof RRT.

Timing of Renal Replacement. Thereis little prospective data regarding theappropriate timing of initiation of RRTand that which are available are incon-clusive. The absolute indications forinitiation of dialysis (severe hyperkale-mia, clinically apparent signs of uremia,severe acidemia, and volume overload, in-cluding pulmonary edema complicatedby hypoxia or cardiogenic shock) arebroadly accepted usual care standards.


Prophylactic dialysis was introduced inthe 1960s (179), and the rst prospectivestudy was published in 1975 comparing aBUN trigger of 70 mg/dL vs. nearly 150mg/dL (180). Survival was 64% in theearly intervention group as comparedto 20% in the non-intensive or standardintervention group (p .01). Conven-tional teaching based on this and otherstudies (181, 182) has been to initiateRRT before a BUN exceeds 100 mg/dL.Unfortunately, not only is the idealBUN not established but also BUN per seis an imperfect reference value because itis widely inuenced by nonrenal factors.

More recently, a review of the datafrom the PICARD study demonstrated anincreased risk of death associated withinitiation of RRT with a BUN 76 mg/dLin comparison to 76 mg/dL (183). Animportant limitation of this study is thatpatients who were conservatively man-aged (did not receive RRT) are invisiblein this analysis, thereby limiting the va-lidity of the ndings regarding impact onmortality. In the only randomized studyof timing of CRRT initiation (n 106),there was no effect on mortality (184).Early dialysis was initiated after 6 hrs ofoliguria. Of the 36 patients included inthe late arm of this study, six patientsdid not receive RRT, of whom four sur-

vived, a fact that likely inuenced theresults of this study. Results from a largeprospective multi-centered observationalstudy of 1200 patients were internallyinconsistent and dependent on the de-nition of early or late initiation ofRRT (185). In this study, late initiationof RRT was associated with worse out-comes (higher crude mortality, longerduration of RRT, increased hospitallength of stay, and greater dialysis depen-dence) when late was dened relative todate of ICU admission. However, therewas no difference in crude mortality if thetiming was dened by serum urea. Fi-nally, there was a lower crude mortality iftiming of RRT initiation was dened bySCr at initiation (higher SCr associatedwith a lower mortality) (185). Unfortu-nately, the question of timing remainsunanswered and controversial (185, 186).There is clearly a need for a large RCT,with a clear denition of early, to helpguide the clinician in determining theappropriate timing for initiation of RRTfor AKI in the ICU.

Choosing a Renal Replacement Dose.Six prospective RCT have been publishedaddressing the question of dose of RRT incritically ill adults (37, 184, 187190; Ta-ble 4). Three of these studies suggest thata higher dose of dialysis translates into

improved outcomes, specically de-creased mortality (37, 187, 188). Ronco etal (187) published the rst RCT in 2000addressing this question. These investiga-tors compared 20, 35, and 45 mL/kg/hrdosing strategies. There was a high mor-tality in all groups but a statisticallylower mortality in the two groups withhigher dose of ultraltration (35 and 45mL/kg/hr) without any difference in com-plication rates between groups (187). In2002, Schif et al (37) found daily dialysisto be superior to alternate day dialysis ina prospective randomized study. Therewere signicantly fewer hypotensive epi-sodes in the daily dialysis group (5% vs.25%). In an intention-to-treat analysis,mortality was 28% for daily dialysis and46% for alternate-day dialysis (p .01)(37). An important limitation of thisstudy is that the delivered dose was sig-nicantly less than the prescribed dose;therefore, the daily dialysis group re-ceived only adequate therapy as judgedby contemporary standards. It may besaid, therefore, that it was a comparisonbetween adequate and inadequate dialy-sis. In 2006, Saudan et al demonstratedthat continuous veno-venous hemodial-tration (CVVHDF); addition of dialysate(11.5 L/hr) to continuous veno-venoushemoltration (12.5 L/hr); improved 28-

Table 4. Summary of randomized controlled trials of dosing strategies for renal replacement therapy for acute kidney injury in the intensive care unit

Author N Design RRT Modality RRT Doses P/D Survival

Randomized controlled trials with mortalitydifferenceRonco et al (187) 425 Single center CVVH (post-lter dilution) (P) 20 mL/kg/hr 15-day: 41%

(P) 35 mL/kg/hr 57%(P) 45 mL/kg/hr 58%

Schif et al (37) 160 Single center Intermittent HD: daily vs.alternate day

Daily HD Kt/V(P) 1.19/(D) 0.92Alternate day HD Kt/V

(P) 1.21/ (D) 0.94

28 day: 72%54%

Saudan et al (188) 206 Single center CVVH vs. CVVHDF(pre-lter dilution)

(D) Mean: 25 mL/kg/hr/87% of prescribed(D) Mean: 42 mL/kg/hr/83% of prescribed

(includes mean 24 mL/kg/hrreplacement and 18 mL/kg/hrdialysate

28-day: 39%59%

Randomized controlled trials without mortality differenceBouman et al (184) 106 Two centers CVVH (post-lter dilution):

early high-volume vs.early low-volume vs. latelow-volume

(D) Mean: 48 ml/kg/hr (early) 28-day: 74%

(D) Mean: 20 ml/kg/hr (early) 69%

(D) Mean: 19 ml/kg/hr (late) 75%

Tolwani et al (189) 200 Single center CVVHDF(pre-lter dilution)

(P) 20 mL/kg/hr/(D) 17 mL/kg/hr(P) 35 mL/kg/hr/(D) 29 mL/kg/hr

ICU discharge or30 day: 56%

49%

Palevsky et al (190) 1124 Multicenter Intensive vs. less intensiveRRT (CVVHDF or SLEDor HD)

(P) 21 mL/kg/hr or SLED or HD 3/wk 60 day: 44%(D) 22 mL/kg/hr or Kt/V 1.3 3/wk 49%(P) 36 mL/kg/hr or SLED or HD 6/wk(D) 35 mL/kg/hr or Kt/V 1.3 6/wk

P, prescribed; D, delivered; CVVH, continuous veno-venous hemoltration; HD, hemodialysis; CVVHDF, continuous veno-venous hemodialtration;SLED, slow low-efciency dialysis.


and 90-day survival compared with he-moltration alone in 206 critically illadults; 39% vs. 59%; p .03 and 34% vs.59%; p .0005, respectively, suggestingthat small solute clearance is important(188).

In contrast, three prospective RCThave demonstrated no difference in mor-tality (184, 189, 190; Table 4). Bouman etal (184), in 2002, showed no difference in28-day mortality when comparing earlyhigh-volume hemoltration, early low-volume hemoltration vs. late low-volume hemoltration with the mediandose (mL/kg/hr) of 48, 20, and 19, respec-tively. More recently, Tolwani et al (189)compared two different doses, 20 mL/kg/hr and 35 mL/kg/hr, of pre-lter CV-VHDF and found no difference in 30-daymortality (44% vs. 51%, p .32). Ofnote, the delivered dose in these twogroups were 17 mL/kg/hr and 29 mL/kg/hr, respectively (189). The largest and onlymulti-centered trial designed to address thequestion of dose of RRT in critically illadults is the acute tubular necrosis studypublished in 2008 (190). This was a two-arm study comparing intensive to standardRRT. The intensive therapy group under-went daily dialysis, CVVHDF, or sustainedlow-efciency dialysis (SLED) at a dose of35 mL/kg/hr, whereas the standard ther-apy group had alternate day dialysis(three times per wk), CVVHDF, or SLEDat 20 mL/kg/hr. Notably, patients wereable to move from intermittent to con-tinuous modalities based on hemody-namic stability but they stayed withintheir assigned intensive or standard treat-ment therapy groups. There was no dif-ference in the primary outcome, deathfrom any cause (190). The RENAL study,comparing CVVHDF 25 mL/kg/hr to 40mL/kg/hr, has completed enrollment butresults have not yet been published.

An important factor in considering theresults of the currently available data arethe difference between study populations,use of solely convective or combinationconvective and diffusive modalities, andthe potential gap between prescribed anddelivered doses. Findings from these neg-ative trials should not be interpreted tomean that dose is not important. On thecontrary, it is likely that dose is impor-tant and, above a minimal dose, furtherescalation may not provide additionalbenet. Based on currently available data,it is our recommendation that to ensurean actual delivered dose of 20 mL/kg/hrfor continuous modalities one must pre-scribe a higher dose (e.g., 25 mL/kg/hr)

to account for lter clotting, time off themachine for interventions, or radio-graphic studies, etc. For intermittentRRT, one should target a Kt/V of 1.2 to1.4 per treatment for alternate day (threetimes per wk) hemodialysis. Further-more, in addition to an appropriate targetdose, there must be close attention givento the actual delivered dose. In summary,one dose does not t all; RRT dose mustbe weight-adjusted.

Choosing a Renal Replacement Mo-dality. Continuous RRT modalities moreclosely approximate normal physiologywith slow correction of metabolic de-rangements and removal of uid. There-fore, CRRT is commonly thought to bebetter-tolerated in the critically ill andhemodynamically unstable patient. Thequestion of superiority remains given theabsence of clear evidence that these ap-parent physiologic advantages translateinto a decrease in ICU or hospital mor-tality (191196).

Since 2000 there have been seven pro-spective RCT designed to address the im-portant clinical question regarding opti-mal RRT modality (192, 193, 195, 197200); of these, only three were multi-centered studies (193, 198, 200). Of note,many of these trials, although publishedafter 2000, enrolled patients in the 1990s.In six of the trials, mortality was theprimary outcome. There have been sev-eral meta-analyses and systematic re-views comparing outcomes of intermit-tent vs. continuous renal replacementmodalities with conicting results (191,201204). A recent meta-analysis (ninerandomized trials) comparing intermit-tent to continuous renal replacementtherapy (intermittent RRT vs. CRRT) inAKI demonstrated no difference in mor-tality or renal recovery (dened as inde-pendence from RRT) (202). Of note, mor-tality was the primary outcome in eightof the nine included trials. Mortality,however, may not be the only clinicallysignicant outcome. Two studies haveshown that CRRT is associated with bet-ter long-term kidney recovery when com-pared to intermittent RRT (205, 206). Incontrast, four RCT that included renalrecovery as a primary outcome showedno difference in need for chronic RRT(193, 195, 198, 200). In the absence ofdenitive data in support of a particularmodality (191, 201), the choice of RRTmodality is currently inuenced by mul-tiple factors, including individual siteavailability, expertise, resources, cost,and likely clinician bias.

Hybrid therapies include SLED andextended daily dialysis. These modalitiesutilize standard intermittent hemodialy-sis machines but provide a slower soluteand uid removal similar to CRRT tech-nologies. Although there have been noprospective randomized trials evaluatingoutcomes, hybrid therapies have beenshown to be safe and effective alternativesto treating AKI in critically ill patients(207, 208).

The question of optimal modality hasnot yet been denitively answered. It isimportant to note that although the datastrongly suggest that there is no differ-ence in outcome between intermittentand continuous modalities, several keypatient populations have been excluded.Namely, hemodynamically unstable pa-tients, brain-injured patients, and thosewith fulminant hepatic failure were ex-cluded and are widely believed to requirecontinuous modalities. Furthermore, acritical limitation of all of the studies isthe absence of a standardized dose (bothwithin and between modalities) (202).RRT, like other medical treatments, mustbe considered in terms of dose adequacyto appropriately draw conclusions regard-ing clinical outcomes. Large randomizedtrials may be necessary to identify otherpotential subsets of patients who mightbenet from continuous modalities.

Anticoagulation is frequently requiredto prevent clotting in extracorporeal cir-cuits. There are no large RCT available toguide the choice of anticoagulation: hep-arin (unfractionated or low-molecular-weight heparin) or citrate-based proto-cols. Bleeding complications remain theprimary concern with anticoagulation.Three small RCT, however, have demon-strated both similar or prolonged lterlife and less bleeding and transfusionwith citrate protocols when compared touse of heparins (209211). In a recentlarger, randomized, non-blinded trial com-paring citrate to nadroparin, circuit sur-vival was similar in both groups, but thecitrate group had a lower mortality rate(212). Currently available data support theuse of citrate for anticoagulation; however,this requires local expertise.

In summary, whereas RRT remainsthe cornerstone of treatment of AKI inthe ICU, many key questions remain con-troversial. This is a rapidly evolving eldand requires early consultation for appro-priate expertise in the management ofRRT for the critically ill patient with AKI.


On the Horizon

The identication of novel candidatebiomarkers of early AKI provides hope forthe success of future clinical early inter-vention trials. Advances in treatment ofAKI have been limited by the inability todiagnose AKI early. Previously failed in-terventions may portend different out-comes if implemented earlier in thecourse of AKI. Novel pharmacologicagents on the horizon include erythro-poietic agents and natriuretic peptides.Novel interventions include the use ofstem cell therapy, renal tubule assist de-vice, and high-ux hemoltration forsepsis.

Candidate Biomarkers. Biomarkers ofAKI in the ICU have three primary poten-tial roles: early detection of AKI, differen-tial diagnosis (e.g., hepatorenal syndromevs. acute tubular necrosis), and prognosis(e.g., need for RRT or mortality). Theideal biomarker for AKI would be sensi-tive, specic, inexpensive, available non-invasively as a point-of-care test, and pro-vide a real-time assessment of GFR. Apanel of biomarkers or kidney functiontests may be needed to address the com-plexity and heterogeneity of AKI in theICU (213). Early identication of AKIwith rapid and reproducible biomarkersis a critical rst step toward improvingoutcomes in AKI.

According to several studies in criti-cally ill patients, serum cystatin C is bet-ter than SCr for early detection of AKI(214, 215) and as a more sensitive markerof small changes in GFR (216218).However, in one smaller study there wasno correlation between cystatin C andSCr (219). In a recent study, urinary cys-tatin C but not plasma cystatin C wassuperior to conventional plasma markersin the early identication of AKI aftercardiac surgery (220). Whereas rapid au-tomated assays for cystatin C are cur-rently available, more information on theuse of cystatin C in the ICU setting and inspecic patient populations (e.g., post-cardiothoracic surgery, sepsis, andtrauma) is necessary before implementa-tion in clinical practice.

Several studies support neutrophil ge-latinase-associated lipocalin (221227),kidney injury molecule-1 (228, 229), andinterleukin (IL)-18 (222, 230, 231) aspromising candidate biomarkers for theearly detection of AKI. Point-of-care testsfor urinary IL-18 and neutrophil gelati-nase-associated lipocalin will likely beavailable for clinical use soon (213, 231

234). Urinary excretion of enzymes (alka-line phosphatase, gamma glutamyltransaminase, N-acetyl-beta-d-glu-cosamine) (235), transporters (sodium-hydrogen exchanger isoform 3) (236), cy-tokines (IL-6, IL-8, and IL-18), andprotein-like substances (fetuin A) (237)are presumably shed into the urinewith AKI; therefore, they may have a rolein the early identication of AKI (232,233).

In addition to emerging biomarkers,promising real-time imaging for use inearly detection of AKI is on the horizon(238, 239). Ongoing discovery using uri-nary proteomic analyses or analysis ofgenetic polymorphisms may identify sus-ceptibility to AKI (240244). Overall, bi-omarkers in AKI, although rapidly evolv-ing, are a eld still in its relative infancy.Their role in the diagnosis and manage-ment of AKI in the ICU, although prom-ising, remains unproven. Furthermore,judging novel biomarkers against an im-perfect gold-standard biomarker (SCr)may have its limitations.

Erythropoietic Agents. The endothe-lium plays a central role in the initiationand maintenance phases of AKI. Animalmodels demonstrate a renal-protective ef-fect of erythropoietin on endotoxin-related kidney injury (245). Decreased se-verity of AKI is proposed to occurthrough tubular regeneration from thedirect effects of erythropoietin on tubularepithelial cells (246). These ndings sup-port the ongoing trials exploring the roleof erythropoietic agents in the preventionor early intervention for AKI using earlybiomarkers (personal communicationand clinicaltrials.gov NCT00476619).

Atrial Natriuretic Peptide. Recombi-nant human atrial natriuretic peptide de-creased the need for dialysis (21% vs.47%) and improved dialysis-free survivalat 21 days (57% vs. 28%) in a RCT of 61complicated post-cardiopulmonary by-pass patients without preexisting CKD(247). Previously, however, in two multi-centered, prospective, randomized trialsin patients with acute tubular necrosis(248) or late oliguric AKI (249), atrialnatriuretic peptide had no effect on needfor dialysis or overall mortality. Furthertrials are needed before the use of atrialnatriuretic peptide can be recommendedfor routine clinical use in cardiac surgerypatients.

Renal Tubule Assist Device. Resultsfrom a recent RCT of the renal tubuleassist device, in which the renal tubuleassist device added to conventional CRRT

was compared to CRRT alone, are prom-ising with respect to both safety and ef-cacy. There was a non-statistically signif-icant decrease in mortality at 28 days anda statistically signicant difference at 180days (secondary outcome) (250).

Hemoltration for Sepsis. Payen et al(251) recently published the ndingsfrom the largest RCT of hemoltrationfor severe sepsis and septic shock. At in-terim analysis, standard CVVH was foundto be deleterious, with increased organfailures in the CVVH group compared tostandard therapy. The study was stoppedat interim analysis and consequently en-rollment was insufcient to detect a dif-ference in mortality with sufcientpower. These ndings contrast with thoseof Honore et al (252) in 2000, suggestinga benecial role for hemoltration in re-fractory septic shock. An important dif-ference between these two studies wasthe delivered dose. In the rst study, thedose, on average, was approximately 2L/hr, whereas in the second study thedose was, on average, 8.7 L/hr for 4 hrs.

Stem Cells and the Kidney. Progenitorcell therapies represent an exciting futureopportunity for treatment of AKI in thecritically ill. Phase 1 trials of mesenchymalstem cells for treatment of patients at highrisk for cardiac surgery-associated AKI areunderway. A phase 2 RCT will be conductedif safety is demonstrated in phase 1 (clini-caltrials.gov ID: NCT00733876).

CONCLUSIONS

Many unanswered questions remainwith respect to early identication, pre-vention, optimal timing, dose, and mo-dality of RRT for AKI in the ICU. Withrespect to AKI in the ICU, the fundamen-tal principal that guides all medical ther-apydo no harmis especially perti-nent. AKI in the ICU most commonlyresults from multiple insults. Therefore,appropriate and early identication of pa-tients at risk for AKI provides an oppor-tunity to prevent subsequent renal in-sults and ultimately impact overall ICUmorbidity and mortality. Strategies toprevent AKI in these patients are of piv-otal importance. Key components of op-timal prevention and management of thecritically ill patient with AKI includemaintenance of renal perfusion andavoidance of nephrotoxins. Whereasmanagement of AKI remains limited pri-marily to supportive care, there are manypotential therapies and interventions onthe horizon.


Although it is widely accepted thatearly intervention therapies have beenlimited by the lack of tools for early de-tection, there are several promising can-didate biomarkers in the pipeline. Fur-thermore, through the establishment ofAKIN, an international and interdiscipli-nary collaborative network with the over-arching objective to address AKI in theICU, there has been tremendous progressin establishing a uniform denition(AKIN criteria) that is valuable for classi-cation, clinical research study design,and prognosis.

A greater appreciation for the role ofAKI in the ICU as an active contributor tomorbidity and mortality is essential tofurthering our knowledge and under-standing of the inuence of AKI in thecritically ill patient. Early detection willfacilitate early intervention. Early inter-vention designed to target the deleterioussystemic effects of AKI will likely improveoverall morbidity and mortality. For now,recognition of risk factors, excellent sup-portive care, and avoidance of clinicalconditions known to cause or worsen AKIremain the cornerstone of managementof AKI in the ICU.

REFERENCES

1. Bagshaw SM, George C, Bellomo R:Changes in the incidence and outcome forearly acute kidney injury in a cohort ofAustralian intensive care units. Crit Care2007; 11:R68

2. Collins AJ, Foley R, Herzog C, et al: Ex-cerpts from the United States Renal DataSystem 2007 annual data report. Am J Kid-ney Dis 2008; 51:S1S320

3. Waikar SS, Wald R, Chertow GM, et al:Validity of International Classication ofDiseases, Ninth Revision, Clinical Modica-tion Codes for Acute Renal Failure. J AmSoc Nephrol 2006; 17:16881694

4. Xue JL, Daniels F, Star RA, et al: Incidenceand mortality of acute renal failure in Medi-care beneciaries, 1992 to 2001. J Am SocNephrol 2006; 17:11351142

5. Chertow GM, Soroko SH, Paganini EP, et al:Mortality after acute renal failure: Modelsfor prognostic stratication and risk adjust-ment. Kidney international 2006; 70:11201126

6. Hoste EA, Clermont G, Kersten A, et al:RIFLE criteria for acute kidney injury areassociated with hospital mortality in criti-cally ill patients: A cohort analysis. CritCare 2006; 10:R73

7. Nash K, Hafeez A, Hou S: Hospital-acquiredrenal insufciency. Am J Kidney Dis 2002;39:930936

8. Mehta RL, Pascual MT, Soroko S, et al:Spectrum of acute renal failure in the in-

tensive care unit: The PICARD experience.Kidney international 2004; 66:16131621

9. Uchino S, Kellum JA, Bellomo R, et al:Acute renal failure in critically ill patients:A multinational, multicenter study. Jama2005; 294:813818

10. Uchino S, Bellomo R, Goldsmith D, et al: Anassessment of the RIFLE criteria for acuterenal failure in hospitalized patients. Criti-cal care medicine 2006; 34:19131917

11. Ostermann M, Chang RW: Acute kidney in-jury in the intensive care unit according toRIFLE. Crit Care Med 2007; 35:18371843;quiz 1852

12. Lin CY, Chen YC, Tsai FC, et al: RIFLEclassication is predictive of short-termprognosis in critically ill patients with acuterenal failure supported by extracorporealmembrane oxygenation. Nephrol DialTransplant 2006; 21:28672873

13. Lopes JA, Jorge S, Resina C, et al: Prognos-tic utility of RIFLE for acute renal failure inpatients with sepsis. Crit Care 2007; 11:408

14. Metnitz PG, Krenn CG, Steltzer H, et al:Effect of acute renal failure requiring renalreplacement therapy on outcome in criti-cally ill patients. Crit Care Med 2002; 30:20512058

15. Liano F, Felipe C, Tenorio MT, et al: Long-term outcome of acute tubular necrosis: Acontribution to its natural history. Kidney-Int 2007; 71:679686

16. Bagshaw SM, Laupland KB, Doig CJ, et al:Prognosis for long-term survival and renalrecovery in critically ill patients with severeacute renal failure: A population-basedstudy. Crit Care 2005; 9:R700R709

17. Chertow GM, Levy EM, HammermeisterKE, et al: Independent association betweenacute renal failure and mortality followingcardiac surgery. Am J Med 1998; 104:343348

18. Chertow GM, Burdick E, Honour M, et al:Acute kidney injury, mortality, length ofstay, and costs in hospitalized patients.J Am Soc Nephrol 2005; 16:33653370

19. Lassnigg A, Schmidlin D, Mouhieddine M,et al: Minimal changes of serum creatininepredict prognosis in patients after cardio-thoracic surgery: A prospective cohortstudy. J Am Soc Nephrol 2004; 15:15971605

20. Waikar SS, Liu KD, Chertow GM: The inci-dence and prognostic signicance of acutekidney injury. Curr Opin Nephrol Hyper-tens 2007; 16:227236

21. Coca SG, Peixoto AJ, Garg AX, et al: Theprognostic importance of a small acute dec-rement in kidney function in hospitalizedpatients: a systematic review and meta-analysis. Am J Kidney Dis 2007; 50:712720

22. Yegenaga I, Hoste E, Van Biesen W, et al:Clinical characteristics of patients develop-ing ARF due to sepsis/systemic inamma-tory response syndrome: Results of a pro-spective study. Am J Kidney Dis 2004; 43:817824

23. Bagshaw SM, Uchino S, Bellomo R, et al:

Septic acute kidney injury in critically illpatients: Clinical characteristics and out-comes. Clin J Am Soc Nephrol 2007;2:431439

24. Bernieh B, Al Hakim M, Boobes Y, et al:Outcome and predictive factors of acute re-nal failure in the intensive care unit. Trans-plant Proc 2004; 36:17841787

25. Vincent JL, de Mendonca A, Cantraine F, etal: Use of the SOFA score to assess theincidence of organ dysfunction/failure in in-tensive care units: Results of a multicenter,prospective study. Working group on sep-sis-related problems of the European Soci-ety of Intensive Care Medicine. Crit CareMed 1998; 26:17931800

26. Harbrecht BG, Rosengart MR, Zenati MS, etal: Dening the contribution of renal dys-function to outcome after traumatic injury.Am Surg 2007; 73:836840

27. Radovic M, Ostric V, Djukanovic L: Validityof prediction scores in acute renal failuredue to polytrauma. Ren Fail 1996; 18:615620

28. Kuitunen A, Vento A, Suojaranta-Ylinen R,et al: Acute renal failure after cardiac sur-gery: evaluation of the RIFLE classication.Ann Thorac Surg 2006; 81:542546

29. Thakar CV, Worley S, Arrigain S, et al: In-uence of renal dysfunction on mortalityafter cardiac surgery: Modifying effect ofpreoperative renal function. Kidney inter-national 2005; 67:11121119

30. Bove T, Calabro MG, Landoni G, et al: Theincidence and risk of acute renal failureafter cardiac surgery. J Cardiothorac VascAnesth 2004; 18:442445

31. Holm C, Horbrand F, von DonnersmarckGH, et al: Acute renal failure in severelyburned patients. Burns 1999; 25:171178

32. Morgera S, Kraft AK, Siebert G, et al: Long-term outcomes in acute renal failure pa-tients treated with continuous renal re-placement therapies. Am J Kidney Dis 2002;40:275279

33. Prescott GJ, Metcalfe W, Baharani J, et al: Aprospective national study of acute renalfailure treated with RRT: Incidence, aetiol-ogy and outcomes. Nephrol Dial Transplant2007; 22:25132519

34. Leacche M, Rawn JD, Mihaljevic T, et al:Outcomes in patients with normal serumcreatinine and with articial renal supportfor acute renal failure developing after cor-onary artery bypass grafting. Am J Cardiol2004; 93:353356

35. Korkeila M, Ruokonen E, Takala J: Costs ofcare, long-term prognosis and quality of lifein patients requiring renal replacementtherapy during intensive care. IntensiveCare Med 2000; 26:18241831

36. Basile DP: Novel approaches in the investi-gation of acute kidney injury. J Am SocNephrol 2007; 18:79

37. Schif H, Lang SM, Fischer R: Daily hemo-dialysis and the outcome of acute renal fail-ure. N Engl J Med 2002; 346:305310

38. Macedo E, Bouchard J, Mehta RL: Renal


recovery following acute kidney injury.Curr Opin Crit Care 2008; 14:660665

39. Mehta RL, Chertow GM: Acute renal failuredenitions and classication: Time forchange? J Am Soc Nephrol 2003; 14:21782187

40. Bellomo R, Ronco C, Kellum JA, et al: Acuterenal failuredenition, outcome mea-sures, animal models, uid therapy and in-formation technology needs: The SecondInternational Consensus Conference of theAcute Dialysis Quality Initiative (ADQI)Group. Crit Care 2004; 8:R204R212

41. Dasta JF, Kane-Gill SL, Durtschi AJ, et al:Costs and outcomes of acute kidney injury(AKI) following cardiac surgery. NephrolDial Transplant 2008

42. Mehta RL, Kellum JA, Shah SV, et al: AcuteKidney Injury Network: Report of an initia-tive to improve outcomes in acute kidneyinjury. Crit Care 2007; 11:R31

43. Bagshaw SM, George C, Bellomo R: A com-parison of the RIFLE and AKIN criteria foracute kidney injury in critically ill patients.Nephrol Dial Transplant 2008; 23:15691574

44. Lopes JA, Fernandes P, Jorge S, et al: Acutekidney injury in intensive care unit pa-tients: A comparison between the RIFLEand the Acute Kidney Injury Network clas-sications. Crit Care 2008; 12:R110

45. Guerin C, Girard R, Selli JM, et al: Initialversus delayed acute renal failure in theintensive care unit. A multicenter prospec-tive epidemiological study. Rhone-AlpesArea Study Group on Acute Renal Failure.Am J Respir Crit Care Med 2000; 161:872879

46. Moran SM, Myers BD: Course of acute renalfailure studied by a model of creatinine ki-netics. Kidney international 1985; 27:928937

47. Tsai JJ, Yeun JY, Kumar VA, et al: Compar-ison and interpretation of urinalysis per-formed by a nephrologist versus a hospital-based clinical laboratory. Am J Kidney Dis2005; 46:820829

48. Perazella MA, Coca SG, Kanbay M, et al:Diagnostic value of urine microscopy fordifferential diagnosis of acute kidney injuryin hospitalized patients. Clin J Am SocNephrol 2008; 3:16151619

49. da Silva Magro MC, de Fatima FernandesVattimo M: Does urinalysis predict acuterenal failure after heart surgery? Ren Fail2004; 26:385392

50. Lee VWS, Harris DCH, Anderson RJ, et al:Acute Renal Failure. In: Diseases of the Kid-ney & Urinary Tract. 8th ed. Schrier RW(Ed). Philadelphia, Wolters Kluwer Health/Lippincott Williams & Wilkins, 2007

51. Miller TR, Anderson RJ, Linas SL, et al:Urinary diagnostic indices in acute renalfailure: a prospective study. Ann Intern Med1978; 89:4750

52. Payen D, de Pont AC, Sakr Y, et al: A posi-tive uid balance is associated with a worse

outcome in patients with acute renal fail-ure. Crit Care 2008; 12:R74

53. Wiedemann HP, Wheeler AP, Bernard GR,et al: Comparison of two uid-managementstrategies in acute lung injury. N EnglJ Med 2006; 354:25642575

54. Ali T, Khan I, Simpson W, et al: Incidenceand outcomes in acute kidney injury: Acomprehensive population-based study.J Am Soc Nephrol 2007; 18:12921298

55. Vivino G, Antonelli M, Moro ML, et al: Riskfactors for acute renal failure in traumapatients. Intensive Care Med 1998; 24:808814

56. De laet I, Malbrain ML, Jadoul JL, et al:Renal implications of increased intra-abdominal pressure: are the kidneys the ca-nary for abdominal hypertension? Acta ClinBelg Suppl 2007:119130

57. Sugrue M, Jones F, Deane SA, et al: Intra-abdominal hypertension is an independentcause of postoperative renal impairment.Arch Surg 1999; 134:10821085

58. Cheatham ML, Malbrain ML, Kirkpatrick A,et al: Results from the International Con-ference of Experts on Intra-abdominal Hy-pertension and Abdominal CompartmentSyndrome. II. Recommendations. IntensiveCare Med 2007; 33:951962

59. Sharp LS, Rozycki GS, Feliciano DV: Rhab-domyolysis and secondary renal failure incritically ill surgical patients. Am J Surg2004; 188:801806

60. Obialo CI, Okonofua EC, Tayade AS, et al:Epidemiology of de novo acute renal failurein hospitalized African Americans: Compar-ing community-acquired vs hospital-ac-quired disease. Arch Intern Med 2000; 160:13091313

61. Wang Y, Cui Z, Fan M: Hospital-acquiredand community-acquired acute renal fail-ure in hospitalized Chinese: A ten-year re-view. Ren Fail 2007; 29:163168

62. Kelleher SP, Robinette JB, Conger JD: Sym-pathetic nervous system in the loss of au-toregulation in acute renal failure. Am JPhysiol 1984; 246:F379F386

63. Schlichtig R, Kramer DJ, Boston JR, et al:Renal O2 consumption during progressivehemorrhage. J Appl Physiol 1991; 70:19571962

64. Bourgoin A, Leone M, Delmas A, et al: In-creasing mean arterial pressure in patientswith septic shock: Effects on oxygen vari-ables and renal function. Crit Care Med2005; 33:780786

65. LeDoux D, Astiz ME, Carpati CM, et al:Effects of perfusion pressure on tissue per-fusion in septic shock. Crit Care Med 2000;28:27292732

66. Friedrich JO, Adhikari N, Herridge MS, etal: Meta-analysis: Low-dose dopamine in-creases urine output but does not preventrenal dysfunction or death. Ann Intern Med2005; 142:510524

67. Kellum JA, M Decker J: Use of dopamine inacute renal failure: A meta-analysis. CritCare Med 2001; 29:15261531

68. Marik PE: Low-dose dopamine: A systematicreview. Intensive Care Med 2002; 28:877883

69. Holmes CL, Walley KR: Bad medicine: Low-dose dopamine in the ICU. Chest 2003; 123:12661275

70. Dunning J, Khasati N, Barnard J: Low dose(renal dose) dopamine in the critically illpatient. Interact Cardiovasc Thorac Surg2004; 3:114117

71. Bellomo R, Chapman M, Finfer S, et al: Low-dose dopamine in patients with early renaldysfunction: A placebo-controlled randomisedtrial. Australian and New Zealand IntensiveCare Society (ANZICS) Clinical Trials Group.Lancet 2000; 356:21392143

72. Lauschke A, Teichgraber UK, Frei U, et al:Low-dose dopamine worsens renal perfu-sion in patients with acute renal failure.Kidney Int 2006; 69:16691674

73. Argalious M, Motta P, Khandwala F, et al:Renal dose dopamine is associated withthe risk of new-onset atrial brillation aftercardiac surgery. Crit Care Med 2005; 33:13271332

74. Devins SS, Miller A, Herndon BL, et al:Effects of dopamine on T-lymphocyte pro-liferative responses and serum prolactinconcentrations in critically ill patients. CritCare Med 1992; 20:16441649

75. Murphy MB, Murray C, Shorten GD:Fenoldopam: A selective peripheral dopam-ine-receptor agonist for the treatment ofsevere hypertension. N Engl J Med 2001;345:15481557

76. Samuels J, Finkel K, Gubert M, et al: Effectof fenoldopam mesylate in critically ill pa-tients at risk for acute renal failure is dosedependent. Ren Fail 2005; 27:101105

77. Brienza N, Malcangi V, Dalno L, et al: Acomparison between fenoldopam and low-dose dopamine in early renal dysfunction ofcritically ill patients. Crit Care Med 2006;34:707714

78. Tumlin JA, Finkel KW, Murray PT, et al:Fenoldopam mesylate in early acute tubularnecrosis: A randomized, double-blind, pla-cebo-controlled clinical trial. Am J KidneyDis 2005; 46:2634

79. Morelli A, Ricci Z, Bellomo R, et al: Prophy-lactic fenoldopam for renal protection insepsis: A randomized, double-blind, place-bo-controlled pilot trial. Crit Care Med2005; 33:24512456

80. Landoni G, Biondi-Zoccai GG, Tumlin JA, etal: Benecial impact of fenoldopam in criti-cally ill patients with or at risk for acute renalfailure: a meta-analysis of randomized clinicaltrials. Am J Kidney Dis 2007; 49:5668

81. Schierhout G, Roberts I: Fluid resuscitationwith colloid or crystalloid solutions in crit-ically ill patients: A systematic review ofrandomised trials. BMJ 1998; 316:961964

82. Alderson P, Schierhout G, Roberts I, et al:Colloids versus crystalloids for uid resus-citation in critically ill patients. CochraneDatabase Syst Rev 2000:CD000567

83. Alderson P, Bunn F, Lefebvre C, et al: Hu-


man albumin solution for resuscitation andvolume expansion in critically ill patients.Cochrane Database Syst Rev 2002:CD001208

84. Alderson P, Bunn F, Lefebvre C, et al: Hu-man albumin solution for resuscitation andvolume expansion in critically ill patients.Cochrane Database Syst Rev 2004:CD001208

85. Choi PT, Yip G, Quinonez LG, et al: Crys-talloids vs. colloids in uid resuscitation: Asystematic review. Critical care medicine1999; 27:200210

86. Finfer S, Bellomo R, Boyce N, et al: A com-parison of albumin and saline for uid re-suscitation in the intensive care unit.N Engl J Med 2004; 350:22472256

87. Myburgh J, Cooper DJ, Finfer S, et al: Salineor albumin for uid resuscitation in pa-tients with traumatic brain injury. N EnglJ Med 2007; 357:874884

88. Schortgen F, Lacherade JC, Bruneel F, et al.Effects of hydroxyethylstarch and gelatin onrenal function in severe sepsis: A multicen-tre randomised study. Lancet 2001; 357:911916

89. Brunkhorst FM, Engel C, Bloos F, et al:Intensive insulin therapy and pentastarchresuscitation in severe sepsis. N Engl J Med2008; 358:125139

90. Cittanova ML, Leblanc I, Legendre C, et al:Effect of hydroxyethylstarch in brain-deadkidney donors on renal function in kidney-transplant recipients. Lancet 1996; 348:16201622

91. Wiedermann CJ: Systematic review of ran-domized clinical trials on the use of hy-droxyethyl starch for uid management insepsis. BMC Emerg Med 2008; 8:1

92. Vincent JL, Sakr Y, Sprung CL, et al: Sepsisin European intensive care units: Results ofthe SOAP study. Crit Care Med 2006; 34:344353

93. Sakr Y, Payen D, Reinhart K, et al: Effects ofhydroxyethyl starch administration on re-nal function in critically ill patients. Br JAnaesth 2007; 98:216224

94. Finfer S, Chittock DR, Su SY, et al: Inten-sive versus conventional glucose control incritically ill patients. N Engl J Med 2009;360:12831297

95. Van den Berghe G, Wilmer A, Hermans G, etal: Intensive insulin therapy in the medicalICU. N Engl J Med 2006; 354:449461

96. van den Berghe G, Wouters P, Weekers F, etal: Intensive insulin therapy in the criticallyill patients. N Engl J Med 2001; 345:13591367

97. Schetz M, Vanhorebeek I, Wouters PJ, et al:Tight blood glucose control is renoprotec-tive in critically ill patients. J Am Soc Neph-rol 2008; 19:571578

98. Lecomte P, Van Vlem B, Coddens J, et al:Tight perioperative glucose control is asso-ciated with a reduction in renal impairmentand renal failure in non-diabetic cardiacsurgical patients. Crit Care 2008; 12:R154

99. Cheung NW, Napier B, Zaccaria C, et al:

Hyperglycemia is associated with adverseoutcomes in patients receiving total paren-teral nutrition. Diabetes Care 2005; 28:23672371

100. Rybak MJ, Abate BJ, Kang SL, et al: Pro-spective evaluation of the effect of an ami-noglycoside dosing regimen on rates of ob-served nephrotoxicity and ototoxicity.Antimicrob Agents Chemother 1999; 43:15491555

101. Prins JM, Buller HR, Kuijper EJ, et al: Onceversus thrice daily gentamicin in patientswith serious infections. Lancet 1993; 341:335339

102. Hatala R, Dinh T, Cook DJ: Once-daily ami-noglycoside dosing in immunocompetentadults: a meta-analysis. Ann Intern Med1996; 124:717725

103. Barza M, Ioannidis JP, Cappelleri JC, et al:Single or multiple daily doses of aminogly-cosides: a meta-analysis. BMJ 1996; 312:338345

104. Munckhof WJ, Grayson ML, Turnidge JD: Ameta-analysis of studies on the safety andefcacy of aminoglycosides given eitheronce daily or as divided doses. J AntimicrobChemother 1996; 37:645663

105. Galloe AM, Graudal N, Christensen HR, etal: Aminoglycosides: Single or multipledaily dosing? A meta-analysis on efcacyand safety. Eur J Clin Pharmacol 1995; 48:3943

106. Ferriols-Lisart R, Alos-Alminana M: Effec-tiveness and safety of once-daily aminogly-cosides: A meta-analysis. Am J Health SystPharm 1996; 53:11411150

107. Harbarth S, Pestotnik SL, Lloyd JF, et al:The epidemiology of nephrotoxicity associ-ated with conventional amphotericin Btherapy. Am J Med 2001; 111:528534

108. Walsh TJ, Finberg RW, Arndt C, et al: Lipo-somal amphotericin B for empirical therapyin patients with persistent fever and neu-tropenia. National Institute of Allergy andInfectious Diseases Mycoses Study Group.N Engl J Med 1999; 340:764771

109. Wingard JR, Wood CA, Sullivan E, et al:Caspofungin versus amphotericin B for can-didemia: A pharmacoeconomic analysis.Clin Ther 2005; 27:960969

110. Mangano DT, Tudor IC, Dietzel C: The riskassociated with aprotinin in cardiac sur-gery. N Engl J Med 2006; 354:353365

111. Huber W, Eckel F, Hennig M, et al: Prophy-laxis of contrast material-induced nephrop-athy in patients in intensive care: Acetylcys-teine, theophylline, or both? A randomizedstudy. Radiology 2006; 239:793804

112. Joannidis M, Schmid M, Wiedermann CJ:Prevention of contrast media-induced ne-phropathy by isotonic sodium bicarbonate:A meta-analysis. Wien Klin Wochenschr2008; 120:742748

113. Ho KM, Morgan DJ: Use of isotonic sodiumbicarbonate to prevent radiocontrast ne-phropathy in patients with mild pre-existing renal impairment: A meta-analysis.Anaesth Intensive Care 2008; 36:646653

114. Hogan SE, LAllier P, Chetcuti S, et al:Current role of sodium bicarbonate-basedpreprocedural hydration for the preventionof contrast-induced acute kidney injury: Ameta-analysis. Am Heart J 2008; 156:414421

115. Meier P, Ko DT, Tamura A, et al: Sodiumbicarbonate-based hydration prevents con-trast-induced nephropathy: A meta-analy-sis. BMC Med 2009; 7:23

116. Navaneethan SD, Singh S, Appasamy S, etal: Sodium bicarbonate therapy for preven-tion of contrast-induced nephropathy: Asystematic review and meta-analysis. Am JKidney Dis 2009; 53:617627

117. Brar SS, Shen AY, Jorgensen MB, et al:Sodium bicarbonate vs sodium chloride forthe prevention of contrast medium-inducednephropathy in patients undergoing coro-nary angiography: A randomized trial. JAMA2008; 300:10381046

118. Tepel M, van der Giet M, Schwarzfeld C, etal: Prevention of radiographic-contrast-agent-induced reductions in renal functionby acetylcysteine. N Engl J Med 2000; 343:180184

119. Kelly AM, Dwamena B, Cronin P, et al:Meta-analysis: Effectiveness of drugs forpreventing contrast-induced nephropathy.Ann Intern Med 2008; 148:284294

120. Gonzales DA, Norsworthy KJ, Kern SJ, et al:A meta-analysis of N-acetylcysteine in con-trast-induced nephrotoxicity: Unsupervisedclustering to resolve heterogeneity. BMCMed 2007; 5:32

121. Hoffmann U, Fischereder M, Kruger B, et al:The value of N-acetylcysteine in the preven-tion of radiocontrast agent-induced ne-phropathy seems questionable. J Am SocNephrol 2004; 15:407410

122. Haase M, Haase-Fielitz A, Ratnaike S, et al:N-Acetylcysteine does not artifactuallylower plasma creatinine concentration.Nephrol Dial Transplant 2008

123. McCullough PA, Bertrand ME, Brinker JA,et al: A meta-analysis of the renal safety ofisosmolar iodixanol compared with low-osmolar contrast media. J Am Coll Cardiol2006; 48:692699

124. Heinrich MC, Haberle L, Muller V, et al:Nephrotoxicity of iso-osmolar iodixanol com-pared with nonionic low-osmolar contrastmedia: Meta-analysis of randomized con-trolled trials. Radiology 2009; 250:6886

125. Holscher B, Heitmeyer C, Fobker M, et al:Predictors for contrast media-induced ne-phropathy and long-term survival: prospec-tively assessed data from the randomizedcontrolled Dialysis-Versus-Diuresis (DVD)trial. Can J Cardiol 2008; 24:845850

126. Cirit M, Toprak O, Yesil M, et al: Angioten-sin-converting enzyme inhibitors as a riskfactor for contrast-induced nephropathy.Nephron Clin Pract 2006; 104:c20c27

127. Onuigbo MA, Onuigbo NT: Does renin-angiotensin aldosterone system blockadeexacerbate contrast-induced nephropathyin patients with chronic kidney disease? A


prospective 50-month Mayo Clinic study.Ren Fail 2008; 30:6772

128. Rosenstock JL, Bruno R, Kim JK, et al: Theeffect of withdrawal of ACE inhibitors orangiotensin receptor blockers prior to cor-onary angiography on the incidence of con-trast-induced nephropathy. Int Urol Neph-rol 2008; 40:749755

129. Marenzi G, Marana I, Lauri G, et al: Theprevention of radiocontrast-agent-inducednephropathy by hemoltration. N EnglJ Med 2003; 349:13331340

130. Marenzi G, Lauri G, Campodonico J, et al:Comparison of two hemoltration protocolsfor prevention of contrast-induced ne-phropathy in high-risk patients. Am J Med2006; 119:155162

131. Cruz DN, Perazella MA, Bellomo R, et al:Extracorporeal blood purication therapiesfor prevention of radiocontrast-induced ne-phropathy: A systematic review. Am J Kid-ney Dis 2006; 48:361371

132. Lee PT, Chou KJ, Liu CP, et al: Renal pro-tection for coronary angiography in ad-vanced renal failure patients by prophylac-tic hemodialysis. A randomized controlledtrial. J Am Coll Cardiol 2007; 50:10151020

133. Tumlin JA, Wang A, Murray PT, et al:Fenoldopam mesylate blocks reductions inrenal plasma ow after radiocontrast dyeinfusion: A pilot trial in the prevention ofcontrast nephropathy. Am Heart J 2002;143:894903

134. Stone GW, McCullough PA, Tumlin JA, etal: Fenoldopam mesylate for the preventionof contrast-induced nephropathy: A ran-domized controlled trial. JAMA 2003; 290:22842291

135. Allaqaband S, Tumuluri R, Malik AM, et al:Prospective randomized study of N-acetyl-cysteine, fenoldopam, and saline for preven-tion of radiocontrast-induced nephropathy.Catheter Cardiovasc Interv 2002; 57:279283

136. Briguori C, Colombo A, Airoldi F, et al:N-Acetylcysteine versus fenoldopam mesy-late to prevent contrast agent-associatednephrotoxicity. J Am Coll Cardiol 2004; 44:762765

137. Bagshaw SM, Ghali WA: Theophylline forprevention of contrast-induced nephropa-thy: A systematic review and meta-analysis.Arch Intern Med 2005; 165:10871093

138. Ix JH, McCulloch CE, Chertow GM: The-ophylline for the prevention of radiocon-trast nephropathy: A meta-analysis. NephrolDial Transplant 2004; 19:27472753

139. Akriviadis E, Botla R, Briggs W, et al: Pentoxi-fylline improves short-term survival in severeacute alcoholic hepatitis: A double-blind, pla-cebo-controlled trial. Gastroenterology 2000;119:16371648

140. Sort P, Navasa M, Arroyo V, et al: Effect ofintravenous albumin on renal impairmentand mortality in patients with cirrhosis andspontaneous bacterial peritonitis. N EnglJ Med 1999; 341:403409

141. Gines P, Tito L, Arroyo V, et al: Randomized

comparative study of therapeutic paracen-tesis with and without intravenous albuminin cirrhosis. Gastroenterology 1988; 94:14931502

142. Gluud LL, Kjaer MS, Christensen E: Terlip-ressin for hepatorenal syndrome. CochraneDatabase Syst Rev 2006:CD005162

143. Ortega R, Gines P, Uriz J, et al: Terlipressintherapy with and without albumin for pa-tients with hepatorenal syndrome: Resultsof a prospective, nonrandomized study.Hepatology 2002; 36:941948

144. Solanki P, Chawla A, Garg R, et al: Bene-cial effects of terlipressin in hepatorenalsyndrome: A prospective, randomized pla-cebo-controlled clinical trial. J Gastroen-terol Hepatol 2003; 18:152156

145. Sanyal AJ, Boyer T, Garcia-Tsao G, et al: Arandomized, prospective, double-blind, pla-cebo-controlled trial of terlipressin for type1 hepatorenal syndrome. Gastroenterology2008; 134:13601368

146. Martin-Llahi M, Pepin MN, Guevara M, et al:Terlipressin and albumin vs albumin in patientswith cirrhosis andhepatorenal syndrome: A ran-domized study. Gastroenterology 2008; 134:13521359

147. Alessandria C, Ottobrelli A, Debernardi-Venon W, et al: Noradrenalin vs terlipressinin patients with hepatorenal syndrome: Aprospective, randomized, unblinded, pilotstudy. J Hepatol 2007; 47:499505

148. Sharma P, Kumar A, Shrama BC, et al. Anopen label, pilot, randomized controlledtrial of noradrenaline versus terlipressin inthe treatment of type 1 hepatorenal syn-drome and predictors of response. Am JGastroenterol 2008; 103:16891697

149. Kiser TH, Fish DN, Obritsch MD, et al:Vasopressin, not octreotide, may be bene-cial in the treatment of hepatorenal syn-drome: A retrospective study. Nephrol DialTransplant 2005; 20:18131820

150. Angeli P, Volpin R, Gerunda G, et al: Rever-sal of type 1 hepatorenal syndrome with theadministration of midodrine and octreotide.Hepatology 1999; 29:16901697

151. Esrailian E, Pantangco ER, Kyulo NL, et al:Octreotide/Midodrine therapy signicantlyimproves renal function and 30-day survivalin patients with type 1 hepatorenal syn-drome. Dig Dis Sci 2007; 52:742748

152. Skagen C, Einstein M, Lucey MR, et al:Combination treatment with octreotide,midodrine, and albumin improves survivalin patients with type 1 and type 2 hepato-renal syndrome. J Clin Gastroenterol 2009;43:680685

153. Cantarovich F, Rangoonwala B, Lorenz H,et al: High-dose furosemide for establishedARF: A prospective, randomized, double-blind, placebo-controlled, multicenter trial.Am J Kidney Dis 2004; 44:402409

154. Sampath S, Moran JL, Graham PL, et al:The efcacy of loop diuretics in acute renalfailure: Assessment using Bayesian evidencesynthesis techniques. Crit Care Med 2007;35:25162524

155. Ho KM, Sheridan DJ: Meta-analysis offrusemide to prevent or treat acute renalfailure. BMJ 2006; 333:420

156. Bagshaw SM, Delaney A, Haase M, et al:Loop diuretics in the management of acuterenal failure: A systematic review and meta-analysis. Crit Care Resusc 2007; 9:6068

157. Mehta RL, Pascual MT, Soroko S, et al:Diuretics, mortality, and nonrecovery of re-nal function in acute renal failure. JAMA2002; 288:25472553

158. Uchino S, Doig GS, Bellomo R, et al: Di-uretics and mortality in acute renal failure.Crit Care Med 2004; 32:16691677

159. Chertow GM, Lazarus JM, Paganini EP, etal: Predictors of mortality and the provisionof dialysis in patients with acute tubularnecrosis. The Auriculin Anaritide Acute Re-nal Failure Study Group. J Am Soc Nephrol1998; 9:692698

160. Venkataram R, Kellum JA: The role of di-uretic agents in the management of acuterenal failure. Contrib Nephrol 2001:158170

161. van der Voort PH, Boerma EC, KoopmansM, et al: Furosemide does not improve renalrecovery after hemoltration for acute renalfailure in critically ill patients: A doubleblind randomized controlled trial. Criticalcare medicine 2009; 37:533538

162. Bagshaw SM, Delaney A, Jones D, et al:Diuretics in the management of acute kid-ney injury: A multinational survey. ContribNephrol 2007; 156:236249

163. McWhirter JP, Pennington CR: Incidenceand recognition of malnutrition in hospital.BMJ 1994; 308:945948

164. Goldstein-Fuchs D: Assessment of nutri-tional status in renal diseases. In: Handbookof Nutrition and Kidney. Mitch WE, Klahr S(Eds). Philadelphia: Lippincott Williams &Wilkins, 2002, pp. 4292

165. Druml W: Nutritional management of acuterenal failure. J Ren Nutr 2005; 15:6370

166. Frankeneld DC, Reynolds HN: Nutri-tional effect of continuous hemodialtra-tion. Nutrition 1995; 11:388393

167. Wooley JA, Btaiche IF, Good KL: Metabolicand nutritional aspects of acute renal fail-ure in critically ill patients requiring con-tinuous renal replacement therapy. NutrClin Pract 2005; 20:176191

168. Cano N, Fiaccadori E, Tesinsky P, et al:ESPEN Guidelines on Enteral Nutrition:Adult renal failure. Clin Nutr 2006; 25:295310

169. Macias WL, Alaka KJ, Murphy MH, et al:Impact of the nutritional regimen on pro-tein catabolism and nitrogen balance in pa-tients with acute renal failure. JPEN J Par-enter Enteral Nutr 1996; 20:5662

170. Bellomo R, Tan HK, Bhonagiri S, et al: Highprotein intake during continuous hemodi-altration: Impact on amino acids and ni-trogen balance. Int J Artif Organs 2002;25:261268

171. Scheinkestel CD, Kar L, Marshall K, et al:Prospective randomized trial to assess ca-


loric and protein needs of critically Ill, an-uric, ventilated patients requiring continu-ous renal replacement therapy. Nutrition2003; 19:909916

172. Fiaccadori E, Maggiore U, Rotelli C, et al:Effects of different energy intakes on nitro-gen balance in patients with acute renalfailure: A pilot study. Nephrol Dial Trans-plant 2005; 20:19761980

173. Ricci Z, Picardo S, Ronco C: Results frominternational questionnaires. ContribNephrol 2007; 156:297303

174. Ricci Z, Ronco C, DAmico G, et al: Practicepatterns in the management of acute renalfailure in the critically

Acute Kidney Injury in Icu

Documents

Transcript of Acute Kidney Injury in Icu