Implementing the Kidney Disease: Improving Global Outcomes/acute kidney injury guidelines in ICU patients

Purpose of reviewAcute kidney injury (AKI) is a frequent finding in critically ill patients and is associated with adverse outcomes. With the purpose of improving outcome of AKI, the Kidney Disease: Improving Global Outcomes (KDIGO) group, a group of experts in critical care nephrology, has presented a set of guidelines in 2012, based on the evidence gathered until mid 2011. This review will update these guidelines with recent evidence. Recent findingsEarly application of a set of therapeutic measures – a bundle – is advised for the prevention and therapy of AKI. Hemodynamic optimization remains the cornerstone of prevention and treatment of AKI. Fluid resuscitation should be with isotonic crystalloids. Recent evidence demonstrated a higher risk for renal replacement therapy (RRT) and mortality in hydroxyethyl starch-exposed patients. Further, blood pressure should be maintained by the use of vasopressors in vasomotor shock. Nephrotoxic drugs should be avoided or stopped when possible. Contrast-associated AKI should be prevented by prehydration with either NaCl 0.9% or a bicarbonate solution. Other therapies, including intravenous N-acetylcysteine and hemofiltration are not recommended. Optimal timing of RRT remains controversial. Fluid overload remains an important determinant for the initiation of RRT. Continuous therapies are preferred in hemodynamically unstable patients; otherwise, choice of modality does not impact on outcomes. SummaryThe KDIGO guidelines as presented in 2012 provide guidelines on the domain of definition of AKI, prevention and treatment, contrast-induced AKI and dialysis interventions for AKI. Especially, early application of a set of measures, the AKI bundle, may prevent AKI and improve outcome.


INTRODUCTION
Acute kidney injury (AKI) is a frequent complication in ICU patients, and is associated with worse outcomes. When defined by the sensitive Risk, Injury, Failure, Loss, End stage kidney disease (RIFLE) definition for AKI or its modifications the Acute Kidney Injury Network (AKIN) or Kidney Disease: Improving Global Outcomes (KDIGO) definitions for AKI, it occurs in one-third to two-thirds of ICU patients [1,2,3 && , [4][5][6][7][8], and approximately 5-10% of ICU patients are treated with renal replacement therapy (RRT) for AKI [9,10,11

DEFINITION OF ACUTE KIDNEY INJURY
An important accomplishment of the last decade was the introduction by the Acute Dialysis Quality Initiative (ADQI) of the RIFLE consensus definition for AKI [1]. It allows comparisons between studies, interventions in AKI patient cohorts with similar severity stage, and as it is also defines very early AKI with low severity, it allows early intervention. This definition was later modified, first by the AKIN [2], and recently by KDIGO [3 && ].
AKI is defined by either an increase of serum creatinine (Scr) or an episode of decreased urine output (Table 1). Importantly, a patient needs to fulfill only one of the criteria for the definition of AKI. Subsequently, the severity of AKI can be graded into one of three severity grades.
There are some issues in this definition that need extra discussion.

Timing
As the emphasis is on acute deterioration of kidney function, the patients should fulfill the criteria within a limited time frame. Therefore, one should compare a new Scr measurement to all Scr measurements in the preceding 7-day period for the 50% increase of Scr, or 48 h for the 0.3 mg/dl Scr increase. If the increase of Scr takes place over a longer period then the patient may be classified as having acute kidney disease. Of note, this time frame is only for the definition of AKI, and is not applicable for staging of the AKI severity grade.

KEY POINTS
AKI is defined and staged by the KDIGO consensus definition and classification system, and prevention and treatment of AKI is best performed by the early application of a 'KDIGO AKI bundle' of measures.
Hemodynamic optimization with isotonic crystalloids and vasopressors is the cornerstone for prevention of AKI.
Consider not using iodinated contrast in risk patients; stop NSAIDs, metformin, diuretics and other nephrotoxic drugs when possible, and prehydrate with NaCl 0.9% or a bicarbonate solution.
The optimal timing of RRT is uncertain, but RRT is advised in life-threatening changes in fluid, electrolyte and acid-base balance.
High dose of RRT offers no benefit over a normal dose, but as the actual dose of RRT is lower than the delivered dose of RRT, the prescription should aim for a bit higher dose, and the dose delivered should be measured. In patients who are in stable condition, ADQI recommended baseline Scr concentration may be obtained within a 3-month period preceding the current event [20].
Clinical judgment is essential for the correct estimation of this baseline Scr. For instance, if Scr is obtained at the end of a preceding ICU admission, it is very unlikely that this value represents the true baseline kidney function. Also, assessment on whether the acute condition of AKI occurred within a 7-day period may be challenging.
When there is no baseline Scr measurement available, the ADQI group advocated the use of the modification of diet in renal disease (MDRD) equation when there is an assumed baseline glomerular filtration rate (GFR) of 75 ml/min or greater [1]. The MDRD equation estimates GFR on sex, age, race and Scr. This method obviously has limitations. The equation was validated in a cohort of US patients, and is therefore not applicable in patients with different body composition such as in Asia or in patients with lower muscle mass, for example, as in critical illness, cirrhosis or paraplegia. Also, it is less precise in patients who have GFR more than 60 ml/min. Despite its shortcomings, this MDRDbased estimation of baseline Scr proved reasonably well in a cohort of ICU patients recruited at three centers in the USA [21]. Alternatively, baseline Scr may be estimated by the use of multiple imputation method [22]. These single center data need to be confirmed in other settings, and the complexity of this method may limit its use in daily practice.

Urine output criteria
The definition requires that urine output is less than 0.5 ml/kg every hour for a 6-h period. This limits its use to ICU patients with a urinary catheter. Studies have used variants of the urine output criteria, for example, urine output less than 3 ml/kg in a period of 6 h, use of fixed blocks of 6 or 8 h similar to the nurses' shift, back calculation of 24-h urine output and so on [23].
There is no indication what patient weight one should use for the oliguria criterion. It seems reasonable to use the 'baseline' patients' weight, as actual patient weight in critically ill patients is seldom measured, and varies according to fluid overload, muscle wasting and weight loss secondary to critical illness. In morbidly obese patients, adjusted body weight may be reasonable although there are no data to support this.

PREVENTION AND TREATMENT OF ACUTE KIDNEY INJURY
This section describes a set of measures that are often described as the 'AKI bundle' (Fig. 1). Summary of the KDIGO recommendations for the prevention and treatment of AKI are as follows: Bundles such as these are attractive and successful as these guarantee that all patients receive care according to the best evidence available. These allow healthcare workers, physicians and nurses to simply tick the measures that need to be done in patients at risk. Implementation of care bundles has proven to improve the outcomes in for instance sepsis [24,25]. Also, several studies in AKI patients have demonstrated that early implementation of simple measures by content experts improved outcomes [14-18,19 && ].

HEMODYNAMIC SUPPORT
Guideline 3.1.1 'In the absence of hemorrhagic shock, we suggest using isotonic crystalloids rather than colloids (albumin or starches) as initial management for expansion of intravascular volume in patients at risk for acute kidney injury or with acute kidney injury (2B). ' As colloids remain in the intravascular compartment for a longer period of time compared with crystalloids, these may seem attractive. However, older hydroxyethyl starch (HES) solutions showed nephrotoxicity and observational data on gelatincontaining solutions suggest that these solutions have similar risk for nephrotoxicity [26,27].
Two well designed studies on modern HES solutions with presumably less nephrotoxicity, further addressed this topic. The 6S trial included 798 critically ill patients with severe sepsis or septic shock and compared a potato-based 6% HES formulation with 130 kD molecular weight (6% HES 130) to Ringer's acetate [28 && ]. At 90 days, patients treated with HES had an increased mortality and a higher prevalence of RRT The guideline also advises against the use of albumin, this because the greater cost of this fluid is not associated with benefit. Albumin was evaluated in the Saline versus Albumin Fluid Evaluation study and was found to be equally effective as saline for fluid resuscitation in ICU patients [33]. Importantly, this was not associated with greater need for RRT. Unfortunately, this study did not report on AKI defined by more sensitive criteria. Albumin-treated patients had a less positive fluid balance compared with saline-treated patients. As a positive fluid balance may impact on the occurrence of AKI and outcome, this may be an argument in favor of the use of albumin [34,35,36 && ,37,38]. It is important to notice that the guideline specifically mentions that isotonic crystalloid solutions should be used for volume resuscitation. Two large studies found that nonisotonic crystalloid solutions containing high concentrations of chloride, such as in NaCl 0.9%, are associated with worse outcomes, including AKI, when compared with crystalloid solutions with lower chloride content such as Plasma-Lyte (Baxter Healthcare Inc., Deerfield, Illinois, USA) [39 && ,40 && ]. These data illustrate that fluids should be seen as drugs with benefit, but also potential toxicity (Fig. 2)   In critically ill patients, systemic hypotension may instigate a decreased renal perfusion eventually leading to AKI. To counter for this, vasopressor therapy is often used in these patients. In a state of vasomotor paralysis, the use of norepinephrine, an a-adrenergic agonist, has beneficial effects on renal blood flow and GFR [42][43][44][45].

Guideline 3.1.3 Protocol-based management of hemodynamic and oxygenation parameters to prevent development or worsening of acute kidney injury in high-risk patients in the perioperative setting (2C) or in patients with septic shock
Goal-directed therapy has been studied extensively in the perioperative setting and in meta-analysis, it was shown that this resulted in less postoperative AKI [46]. Although the beneficial effect of early goaldirected therapy (EGDT) in ICU patients on prevention of AKI is plausible, the evidence is still limited. The landmark study by Rivers et al. [47] on EGDT for severe sepsis unfortunately reported no data on AKI. In another study on EGDT in septic ICU patients by Lin et al. [48], EGDT patients had a lower incidence of AKI compared with controls (38.9 vs. 55.2%, P ¼ 0.015). At present, several large multicenter studies on EGDT are under way that will provide additional evidence for its use.

GLYCEMIC CONTROL AND NUTRITIONAL SUPPORT
AKI patients are severely catabolic, and nutritional support is, therefore, an important aspect in the therapeutic plan for these patients. KDIGO summarized the available evidence on this topic and could only formulate recommendations with 2C or 2D grade of evidence, that is suggestions, with low or very low quality of evidence, mostly based upon expert opinion.  The landmark study of Van den Berghe et al. [49] introduced the concept of intensive insulin therapy (IIT) in the ICU. In surgical -and a subgroup of medical -ICU patients, IIT protocol improved outcome and lowered the incidence of AKI [50,51]. These beneficial results could not be reproduced in subsequent studies on IIT [26,52]. One of the great concerns about IIT is the occurrence of hypoglycemia and its impact on outcome [53]. The KDIGO group suggests the use of a less stringent insulin therapy protocol targeting plasma glucose 110-149 mg/dl in critically ill patients.
On the basis of guidelines by expert panels, a total energy intake of 20-30 kcal/kg per day is suggested, preferably via enteral route. Proteins should not be restricted with the aim for preventing or delaying RRT. It is suggested to administer 0.8-1.0 g/kg per day in AKI patients not treated with RRT, and 1.0-1.5 g/kg per day in patients treated with RRT, up to 1.7 g/kg per day in patients treated with CRRT.

Guideline 3.4 We recommend not using diuretics to prevent or treat acute kidney injury, except in the management of fluid overload
The available evidence from small studies cannot demonstrate that AKI is prevented with the use of diuretics, or that AKI patients have faster recovery [54,55]. So far, diuretics only have a role in the management of volume overload. Guideline 3.5 and 3.7: It is not recommended to use low-dose dopamine to prevent or treat acute kidney injury, similarly it is suggested not to use fenoldopam or atrial natriuretic peptide to prevent or treat acute kidney injury Although vasodilation and increasing renal blood flow may seem a logic therapy for the prevention and treatment of AKI, this has not been proven in studies. The evidence for being not beneficial is strongest for low-dose dopamine [56][57][58].

Guideline 3.6: Growth factor intervention
Three observational studies in cardiac surgery found that erythropoietin (EPO)-treated patients prevented AKI [59][60][61]. However, these results could not be confirmed in an early intervention study in ICU patients and in cardiac surgery patients [62 && ,63]. KDIGO recommends, therefore, evaluating the usefulness of EPO in randomized controlled trials (RCTs).

Guideline 3.8 Prevention of aminoglycosideand amphotericin-related acute kidney injury
Given the nephrotoxicity of aminoglycosides and amphotericin, it is suggested to limit their use to infections in which no alternative antimicrobial drug is available. Aminoglycosides should be administered preferably once daily, and drug levels should be monitored daily.
Amphotericin should be given as a lipid formulation in order to reduce the nephrotoxicity.

CONTRAST-INDUCED ACUTE KIDNEY INJURY
Contrast media cause nephrotoxicity, but other risk factors for the development of AKI are often present in the critically ill patient. For that reason, the term contrast-associated AKI (CA-AKI) may seem more appropriate [64]. CA-AKI occurs in 10-22.5% of ICU patients, seldom requires RRT, and is associated with mortality, even on long-term follow-up [64,65 & , [66][67][68]. ICU patients should be assessed for the risk of CA-AKI (preexisting renal impairment, diabetes, nephrotoxic agents, advanced age, hemodynamic instability or hypertension). One should always consider not administering iodinated contrast. The lowest possible dose of modern low or isoosmolar contrast agents should be used. Ideally, NSAIDs, metformin and diuretics are stopped 1 day beforehand. In patients who are at risk for CA-AKI, intravenous volume expansion is recommended, either by administering saline (NaCl 0.9%) or a bicarbonate solution (846 ml glucose 5% þ 154 ml of 1000 mEg/l NaHCO3) at a rate of 3 ml/kg for 1 h before and 1 ml/kg per hour for 6 h after contrast administration [69-71,72 && ]. Although meta-analyses suggest benefit for the bicarbonate solution over saline, bias and heterogeneity limit this recommendation. In case the bicarbonate solution needs to be prepared, this may be associated with errors.
As the data on prevention of CA-AKI by NAC are conflicting, IV NAC is not recommended. But, given its beneficial potential and low toxicity, oral NAC should be administered in patients who are at risk for developing AKI. For most ICU patients, this seems less applicable, as most studies on oral NAC were performed in patients undergoing elective coronary angiography, with administration the night before contrast.
Evidence for the administration of fenoldopam or theophylline in patients who are at risk for AKI is lacking. Similarly, data supporting the prophylactic use of RRT in patients at increased risk for CA-AKI are insufficient.

DIALYSIS INTERVENTIONS FOR TREATMENT OF ACUTE KIDNEY INJURY
Although RRT has been in use for more than half a century, many aspects of this therapy remain controversial.

Timing and initiation of renal replacement therapy
It seems plausible that early initiation of RRT may positively impact on outcomes in ICU patients with AKI. But because of the possible side-effects of this invasive therapy (hypotension, arrhythmia, hemorrhage and complications of vascular access), there is a tendency to avoid RRT as long as possible. Also, RRT-induced hypoperfusion of the kidneys may impair kidney recovery and increase the progression of CKD [73].
Because of the lack of evidence, the KDIGO recommendations concerning timing of RRT in AKI are not graded. Initiation of RRT is advised in life-threatening changes in fluid, electrolyte and acid-base balance.
Extracorporeal therapy can either function as renal replacement (when no kidney function is present) or renal support RRT (as an adjunct to kidney function). The following list shows potential applications for RRT: Historically, timing of initiation of RRT was based on serum urea. However, serum urea is determined by many other variables that have no relation to kidney function [74]. In addition, recent studies could not demonstrate that urea differentiates between outcomes [75][76][77].
Metabolic acidosis is a complication that frequently occurs in ICU patients with AKI, but initiation of RRT in ICU patients with AKI and metabolic acidosis is still a matter of debate. As RRT does not treat the underlying cause of the acidosis, it can only provide restoration of homeostatic equilibrium and fluid balance, enabling specific therapeutic measures [78].
Numerous observational studies indicate fluid overload as an important determinant of worse outcomes [34,35,36 && ,37,38]. Further, a subanalysis from the Randomized Evaluation of Normal versus Augmented Level Replacement Therapy (RENAL) study showed that a negative fluid balance during CRRT was associated with better survival [79 & ]. Therefore, fluid overload may be an important determinant for initiation of RRT. But also on this topic, prospective studies that randomized initiation of RRT based upon fluid status are absent.

Criteria for stopping renal replacement therapy
In literature, data considering the decision to stop RRT are even scarcer. Therefore, KDIGO issues a pragmatic and nongraded recommendation that RRT should be discontinued when kidney function has recovered.

Anticoagulation
Patients without increased bleeding risk on intermittent RRT are recommended anticoagulation with unfractionated or low-molecular-weight heparin.
When CRRT is used, citrate anticoagulation is recommended, unless there are contra-indications for citrate such as reduced liver function or shock with reduced muscle perfusion. Data from five randomized studies showed that citrate-based protocols were associated with longer filter life, less bleeding and in one study also better survival [80][81][82][83][84].
Regional heparin anticoagulation, in which unfractionated heparin is neutralized after the filter with protamine is not advised. This is because the longer half-life of heparin makes it extremely difficult to titrate.
In patients with heparin-induced thrombocytopenia, heparin must be stopped, and thrombin inhibitors such as argatroban or factor Xa inhibitors (danaparoid or fondaparinux) are recommended.
Vascular access for renal replacement therapy in acute kidney injury As in end stage renal disease patients, central vein stenosis is more frequently seen in subclavian dialysis catheters [85,86], KDIGO recommends the right jugular vein, followed by the femoral vein as the optimal insertion place.

Modality of renal replacement therapy in acute kidney injury patients
Generally, the choice of modality of RRT is based on the availability of a specific modality or local experiences. A Cochrane Collaboration meta-analysis including RCTs that compared CRRT with intermittent hemodialysis (IHD) in AKI patients, could not demonstrate differences in hospital and ICU mortality, length of hospital stay or renal recovery [87,88 && ]. CRRT and IHD should, therefore, be seen as complementary therapies except for patients with AKI who are hemodynamically unstable or present with increased intracranial pressure. In these cases, CRRT is considered the optimal modality of RRT.

Dose of renal replacement therapy in patients with acute kidney injury
The concept of dialysis dose is frequently addressed in literature. However, the available evidence is limited and conflicting because of the differences in study design and poor quality of reporting data [89]. Two recently published trials have assessed dialysis dose in critically ill patients with AKI. Both the RENAL and Acute Renal Failure Trial Network trial compared high dose vs. normal dose RRT and could not demonstrate differences in mortality or renal recovery [90,91]. On the basis of these data, KDIGO recommends in IHD and extended dialysis to deliver a weekly Kt/V of 3.9. For CRRT, an effluent volume of 20-25 ml/kg per min is recommended. Because of downtime, this will require a higher prescription.

CONCLUSION
The guidelines proposed by KDIGO propose an extensive overview of the current state of the art for AKI. The RIFLE and AKIN definitions for AKI have been modified into an updated version: the KDIGO definition and grading system. Similar to, for example, sepsis, early application of a bundle of measures is proposed for the prevention of AKI: the 'KDIGO AKI bundle'. These include avoidance of nephrotoxic agents, optimizing hemodynamic status, guidelines for the prevention of CA-AKI and guidelines for the processes of care for RRT.