Kidney and patient outcomes after acute kidney injury in adults

INTRODUCTION — The term "acute kidney injury" (AKI) has replaced the more traditional term acute renal failure, which encompassed all conditions with a sudden loss of excretory kidney function and was not related to a specific mechanism of injury [1,2].

Clinical and epidemiologic studies that discuss prolonged or severe AKI, particularly in the critical care setting, are predominantly focused on intrinsic AKI or acute tubular necrosis, but it is important to note that the latter is not the only form of AKI.

Kidney and patient outcomes following an episode of AKI are reviewed here [3-5].

The diagnosis, pathogenesis, and possible treatment of this disorder and issues related to the definition of AKI are discussed separately:

●(See "Diagnostic approach to adult patients with subacute kidney injury in an outpatient setting".)

●(See "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults".)

●(See "Pathogenesis and etiology of ischemic acute tubular necrosis".)

●(See "Possible prevention and therapy of ischemic acute tubular necrosis".)

●(See "Definition and staging criteria of acute kidney injury in adults".)

EPIDEMIOLOGY — The incidence and prevalence of AKI are not well known, which is due in part to differences in definition and characteristics of assessed patients. (See "Definition and staging criteria of acute kidney injury in adults".)

In a meta-analysis that included 143 studies and >3.5 million patients, mostly from hospital settings, the global incidence of AKI using the Kidney Disease: Improving Global Outcomes (KDIGO) definition was 22 percent [6]. In another study of 1802 patients admitted to one of 97 intensive care units (ICUs) in 33 countries, the incidence of AKI using the KDIGO definition was as high as 57 percent [7]. The KDIGO definition of AKI is discussed elsewhere. (See "Definition and staging criteria of acute kidney injury in adults".)

The incidence of AKI appears to be increasing in the United States [8-11]. As an example, in one study, the incidence of AKI rose from 61 per 100,000 population in 1988 to 288 per 100,000 population in 2002 [8]. The incidence of dialysis-requiring AKI increased sevenfold between 1988 and 2002 [8] and continued to rise by 10 percent per year until 2009 [9]. Similarly, between 2000 and 2014, AKI-related hospitalizations increased by 140 percent in diabetic and 230 percent in nondiabetic individuals [10].

Some possible factors contributing to the rise in the incidence of AKI include the following [12,13]:

●Aging of the population

●Rising incidence of comorbidities that affect susceptibility to AKI, such as diabetes mellitus, hypertension, heart failure, chronic kidney disease, cancer, and sepsis

●Increasing clinician awareness about AKI

●Use of more sensitive definitions for the diagnosis of AKI (leading to inclusion of less severe AKI)

●Increased use of nephrotoxins such as newer chemotherapeutic agents and antimicrobial agents

●Increasing frequency of invasive and surgical procedures

The epidemiology of and risk factors for AKI occurring after coronary artery bypass graft surgery are discussed separately. (See "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Acute kidney injury'.)

DETERMINANTS OF KIDNEY OUTCOMES

Duration of acute kidney injury — Patients with AKI have a kidney failure phase that typically lasts between 7 and 21 days [3]. However, the duration of AKI is variable, depending upon the length and severity of the initial episode; whether or not recurrent insults continue; and, perhaps, whether the patient is oliguric or nonoliguric [3]. Some patients recover within days, while others require dialysis for many months. (See "Nonoliguric versus oliguric acute kidney injury".)

The importance of AKI duration for overall and cardiovascular prognosis was examined in a meta-analysis of 18 retrospective studies [14]. Compared with those who did not have AKI, those with long-duration AKI (defined as lasting seven or more days) had the highest mortality risk (relative risk [RR] 2.28, 95% CI 1.77-2.94), whereas the increase in risk was smaller among those with short-duration AKI (defined as lasting two or fewer days; RR 1.42, 95% CI 1.21-1.66). Duration of AKI was also associated with adverse cardiovascular outcomes and development of chronic kidney impairment [14].

Rapid recovery — A short duration of AKI has been described in patients undergoing suprarenal aortic clamping for aortic aneurysm surgery, resulting in 20 to 80 minutes of total kidney ischemia. In such situations, even though clinical evidence of AKI can be present postoperatively (low glomerular filtration rate [GFR], elevated fractional excretion of sodium), kidney function usually returns to baseline within one to three days (figure 1) [3,15].

Although its pathophysiology differs from that of ischemic AKI, the limited duration of the insult may also explain, in part, the typical three- to five-day course of contrast-associated AKI. (See "Contrast-associated and contrast-induced acute kidney injury: Clinical features, diagnosis, and management".)

Two potential mechanisms for a short duration of kidney dysfunction in patients with AKI include mild injury and postischemic tubular dysfunction. Postischemic tubular dysfunction is a phenomenon in which the tubular cells remain morphologically normal but are functionally impaired in a manner similar to a postischemic "stunned" myocardium.

Tubular sodium reabsorption occurs via entry into the cells across the luminal membrane and then active transport out of the cells by the Na-K-ATPase pump in the basolateral membrane. In animal models, following kidney ischemia, the actin cytoskeleton that anchors the Na-K-ATPase pump to the basolateral membrane becomes disrupted, resulting in partial redistribution of the pumps onto the luminal membrane, thereby resulting in luminal sodium loss [16,17]. Recovery of function is associated with the return of the Na-K-ATPase pumps to the basolateral membrane [17].

Prolonged acute kidney injury — Recovery of kidney function may be markedly delayed in patients with persistent septic physiology and circulatory shock or who have repeated episodes of kidney ischemia and/or repeated exposure to endogenous or exogenous nephrotoxins [3,18]. In these settings, tubular cell apoptosis and/or necrosis have occurred, and the regeneration of tubular cells that is required for the restoration of kidney function is delayed or completely abolished.

In addition, hemodynamic autoregulation in kidneys can be altered, creating an environment that is particularly sensitive to diminished perfusion. Normal kidneys vasodilate in the presence of ischemia to maintain renal blood flow as part of the normal autoregulatory response. This response is impaired in AKI, perhaps because vascular endothelial injury hampers the release of vasodilating substances such as prostacyclin and nitric oxide (endothelium-derived relaxing factor) [19].

As a result, recurrent episodes of hypotension, as in sepsis or with kidney replacement therapy (KRT), are more likely to result in recurrent tubular injury. KRT may also delay recovery of kidney function by neutrophil activation resulting from blood contact with the dialyzer membranes. (See "Evaluation and management of suspected sepsis and septic shock in adults" and "Dialysis-related factors that may influence recovery of kidney function in acute kidney injury (acute renal failure)", section on 'Characteristics of the dialysis membrane'.)

Nevertheless, even patients with prolonged disease who require KRT for eight weeks or more will, if they survive and leave the hospital, usually recover sufficient function to allow KRT to be discontinued [18,20,21]. (See "Dialysis-related factors that may influence recovery of kidney function in acute kidney injury (acute renal failure)".)

Degree of recovery — Multiple studies have demonstrated an increase in the risk of chronic kidney disease (CKD) and end-stage kidney disease (ESKD) among patients who recover from in-hospital AKI [22-27]. This risk of CKD is high even among patients with no prior history of kidney disease and who recover completely from the episode of AKI [25,27]. Patients who recover from dialysis-requiring AKI often have incomplete recovery and are left with stage 4 or 5 CKD [22,28].

Risk factors for development of CKD or ESKD after AKI include albuminuria, lower baseline estimated GFR (eGFR), older age, female sex, greater severity of AKI, higher creatinine at discharge, and presence of multiple comorbidities such as heart failure, diabetes, or hypertension [26,29,30]. In one study, presence of albuminuria at three months was the strongest predictor of future CKD and ESKD [26]. However, in that study, albuminuria had not been ascertained prior to AKI, and, therefore, it was unclear if albuminuria resulted from the AKI or if it represented a pre-existing risk factor for future CKD.

Patients with pre-existing CKD who develop AKI are more likely to progress to ESKD compared with those who never develop AKI [31]. In addition, patients with pre-existing CKD are at a higher risk, compared with patients without pre-existing CKD, of requiring dialysis when they develop AKI [32,33], and of remaining on dialysis after hospital discharge [24,34-36].

Patients who recover from in-hospital AKI should be evaluated at three months (or sooner, depending upon the severity of AKI and the degree of recovery) to monitor for resolution of AKI, for development of new-onset CKD, or for the worsening of pre-existing CKD [37].

Recurrent AKI — Recurrent AKI is defined as repeated AKI occurring after partial or complete recovery of, and within 12 months of, an initial (index) episode of AKI [38-42]. In various studies of inpatients, recurrent AKI occurred in 25 to 31 percent within six months after the index episode and was associated with an increased risk of CKD and death, compared with patients who did not develop recurrent AKI.

Risk factors for recurrent AKI consistently identified across multiple studies include older age, comorbidities such as HF, coronary artery disease (CAD), diabetes, liver disease, cancer, and CKD at baseline [41,42]. Some studies have also reported an association between the severity of the index AKI episode and recurrent AKI [38,41]. These data highlight the importance of close follow-up for those with AKI, as well as clearly and consistently documenting clinical events and course related to AKI in hospital discharge summaries in order to improve the transition of care between hospital and ambulatory settings. As an example, in a cross-sectional study of randomly selected hospitalized patients with AKI, less than one-half of clinician discharge summaries and an even smaller percentage of patient discharge instructions documented the presence, cause, or course of AKI during hospitalization [43].

The best data come from a study that included 38,659 patients hospitalized for AKI within a single hospital system and found that recurrent AKI occurred in 29 percent of patients at a median of six months following the index episode [42]. Recurrent AKI was associated with an increased rate of death over a median follow-up of 1.8 years (adjusted hazard ratio [aHR] 1.66, 95% CI 1.56-1.77). Black patients and Hispanic patients had a higher risk; other risk factors identified were older age, lower body mass index (BMI), proteinuria, anemia, baseline CKD, ischemic stroke, diabetes, chronic liver disease, chronic lung disease, hypothyroidism, systemic cancer, and hemorrhage.

A study from the VA hospital system included 11,683 hospitalizations with AKI and found that 25 percent of patients had recurrent AKI at a median of two months [41]. Patients with recurrent AKI had a higher one-year mortality compared with patients who did not have recurrent AKI (35 versus 18 percent). Comorbid conditions such as HF, CAD, advanced liver disease, cancer, volume depletion, baseline CKD, dementia, hypoalbuminemia, HIV, and diabetes were identified as risk factors. Severity of the index AKI event was also associated with recurrence.

Recurrent episodes of AKI are associated with an increased risk of developing CKD. As an example, a study from the VA hospital system that included 530 diabetic patients admitted for AKI found that 30 percent of patients had recurrent AKI on follow-up [40]. Additionally, they found that each episode of AKI (counting up to three events for each patient) doubled the risk of occurrence of stage 4 CKD (aHR 2.02, 95% CI 1.78-2.32), regardless of the baseline eGFR. However, it remains unclear if recurrent AKI itself results in progression to CKD, or if it reflects other factors that predispose to both recurrent AKI and progression of CKD.

Kidney replacement therapies — The effects of KRT on recovery of kidney function are controversial. The choice of initial KRT modality (continuous versus intermittent) does not affect the likelihood of recovery of kidney function [44].

Individual high-quality trials failed to show an association between intensity of KRT and recovery of kidney function [45,46]. However, a meta-analysis of eight trials (that included high- and low-quality studies) showed delayed kidney function recovery with higher intensity therapy (30 versus 25 percent remaining dialysis dependent at 28 days) [47]. Intensity of KRT in these trials was defined either by a higher frequency of intermittent dialysis (six versus three times per week) or by a higher dose of continuous venovenous hemofiltration (35 to 40 versus 20 to 25 mL/kg/hour).

The use of biocompatible membranes and high-flux membranes for KRT in patients with AKI might play a role in recovery of kidney function, although the data are limited and inconsistent [48]. Further details on KRT in AKI are discussed separately. (See "Kidney replacement therapy (dialysis) in acute kidney injury in adults: Indications, timing, and dialysis dose".)

PATIENT SURVIVAL — AKI during a hospitalization is associated with high in-hospital and long-term mortality.

Among patients in the intensive care unit (ICU) who develop AKI requiring dialysis, reported mortality rates range from 40 to >60 percent [21,49-56]. In-hospital mortality rates are lower among patients with AKI in the general hospital population or who have less severe AKI (15 to 30 percent versus 1 to <10 percent among patients without AKI) [8,57-61]. A small, acute decrement in kidney function due to AKI may also increase in-hospital mortality [62].

Post-hospitalization mortality of patients with AKI is also increased [25,59,61,63-67]. The following observations illustrate the range of findings:

●In one study, among 36,980 patients admitted to a Veterans Affairs (VA) facility between 1999 and 2005, compared with patients admitted with myocardial infarction (MI) alone, those admitted with AKI or AKI plus MI had a higher risk of death after discharge (hazard ratios [HRs] 1.85, 95% CI 1.76-1.94 and 2.14, 95% CI 2.05-2.23, respectively) [68].

●In a large meta-analysis of 110 studies (429,535 patients) that utilized the Kidney Disease: Improving Global Outcomes (KDIGO) definition of AKI, AKI-associated mortality was 23 percent [6].

●In a study of over five million hospital discharges, death within 90 days after hospital discharge was 35 and 13 percent among patients with and without AKI, respectively [61]. The major causes of in-hospital mortality in patients with AKI are infection and the underlying disease (such as hypotension following open-heart surgery), not necessarily kidney failure itself. Patients at risk are generally very ill, with evidence of multiple organ dysfunction.

●Longer-term mortality among patients with AKI may also be increased, as demonstrated in a study of 843 patients who underwent cardiac surgery [59]. Eight years after hospital discharge, mortality was significantly higher in patients with postoperative AKI (defined as at least a 25 percent elevation in serum creatinine in the first postoperative week) compared with those who did not have AKI (36 versus 22 percent; adjusted HR 1.63, 95% CI 1.15-2.32). The increase in risk was independent of whether or not kidney function had recovered at the time of hospital discharge. However, the increase in late mortality did not begin until more than four to five years after surgery, raising a question about the relationship to postoperative AKI. (See "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Acute kidney injury'.)

●In a systematic review of 47,017 patients that examined the long-term morbidity and mortality risk among survivors of AKI, the incidence rate of mortality was 8.9 deaths per 100 person-years in survivors of AKI and 4.3 deaths per 100 patient-years in survivors without AKI (rate ratio [RR] 2.59, 95% CI 1.97-3.42) [66]. AKI was also associated with development of MI. The incidence rate of chronic kidney disease (CKD) after an episode of AKI was 7.8 events per 100 patient-years, and the rate of end-stage kidney disease (ESKD) was 4.9 events per 100 patient-years. Unfortunately, the relative risk (RR) of development of CKD after episodes of AKI was not determined, because of lack of follow-up of control subjects who did not have AKI. The majority of patients who sustain an episode of AKI are not followed by nephrologists after discharge from the hospital [69]. This is confirmed by a survey of 598 United States nephrologists who stated that only 9 percent of outpatient clinic visits were for follow-up of AKI [70].

Factors influencing mortality — An understanding of the clinical characteristics that appear to most strongly affect mortality among patients with AKI may help direct supportive care to those most likely to benefit. A variety of factors have been associated with increased mortality, including male sex, race, older age, oliguria, cardiovascular and cerebrovascular events, overall severity of illness, and quality of life after recovery from AKI [67,71-77]. A multicenter, prospective cohort study evaluated the natural history and predictors of mortality in 618 critically ill patients with AKI, 64 percent of whom required kidney replacement therapy (KRT) [76]. Multivariable logistic regression analysis revealed that the following characteristics were predictive of mortality at 60 days:

●Age (odds ratio [OR] 1.13 per decade)

●Sepsis (OR 1.50)

●Adult respiratory distress syndrome (OR 1.79)

●Liver failure (OR 1.62)

●Thrombocytopenia (OR 1.66)

●Blood urea nitrogen (BUN; OR 1.09 per 10 mg/dL [3.6 mmol/L])

●Serum creatinine <2.0 mg/dL (OR 1.99 [177 micromol/L])

The underlying importance of the overall severity of illness in determining survival among patients with AKI was supported by the findings of a multicenter, prospective, controlled trial of 153 patients with AKI requiring KRT [78]. Illness severity within 24 hours of the initiation of KRT, as determined by the Acute Physiologic and Chronic Health Evaluation (APACHE) II scoring system, predicted survival and recovery of kidney function. Other observational studies also showed an association of the APACHE II and other severity of illness scores with hospital mortality [79,80]. (See "Predictive scoring systems in the intensive care unit".)

Acute pancreatitis represents a good example of the relation between multiple organ dysfunction and patient survival in AKI. A report of 267 patients with acute pancreatitis noted a 16 percent incidence of AKI [81]. Kidney failure tended to occur relatively late in the course, usually in the setting of multiple organ failure. The mortality rate in these patients was 81 percent. (See "Clinical manifestations and diagnosis of acute pancreatitis".)

In contrast, the prognosis is usually excellent in patients who are otherwise healthy, as with AKI following a transfusion reaction, the administration of a nephrotoxic drug, or ischemia in which perfusion has been restored. This was illustrated by the results among patients enrolled in the placebo arm of a multicenter, randomized trial; the mortality rate was 10 versus 30 percent among patients with nephrotoxic versus ischemic acute tubular necrosis, respectively [82].

Pre-existing poor nutritional status may enhance mortality in AKI. In a prospective study of 309 patients with AKI, a significantly higher proportion of patients with severe malnutrition died compared with those with a normal nutritional status (62 versus 13 percent, respectively) [83]. Similar findings were noted in a second series, in which both hypoalbuminemia and hypocholesterolemia, each assumed to reflect poor nutritional status, were associated with a significantly increased risk of death among patients with AKI [84].

Race may also influence the effect of AKI on in-hospital mortality. One study of approximately 400,000 patients with AKI found a significantly lower in-house mortality among Black patients compared with White patients [71]. This was observed for almost all diagnoses and procedures. As an example, a significantly lower mortality rate was noted among Black patients with AKI post-coronary artery bypass surgery (OR 0.73, 95% CI 0.62-0.87). The reasons underlying this observation are not clear.

Health-related quality of life may predict mortality among survivors of AKI. This was suggested by a multivariate analysis of 429 AKI survivors who were identified from a randomized, multicenter study [77]. A low health-related quality of life, assessed at day 60 using the Health Utilities Index Mark 3 (HUI3), was associated with a higher mortality at one year compared with higher scores. Four prespecified HUI3 attributes, including higher ambulation, emotion, cognition, and subscale pain scores, were associated with better survival. Other variables that were associated with an increased risk of death in this study included older age, comorbid disease, and a longer length of initial hospital stay. Dialysis dependence at day 60 was not associated with increased mortality on the multivariate analysis.

Kidney replacement therapy factors and mortality — Patient survival is not different among patients dialyzed using either a biocompatible or a bioincompatible membrane [48].

The effects of timing of initiation and modality of kidney replacement therapy on patient survival have been controversial. These issues are discussed separately. (See "Dialysis-related factors that may influence recovery of kidney function in acute kidney injury (acute renal failure)".)

Does kidney failure itself affect prognosis? — A separate question is whether the development of AKI itself directly contributes to mortality. Possible mechanisms for such an effect are decreased ability to fight infection and altered platelet function with kidney failure, changes that could predispose to sepsis and bleeding, as well as the release of mediators that contribute directly to acute lung injury [85].

The possible independent association of AKI with increased mortality in acutely ill hospitalized patients is illustrated by the following studies [56,59,60,86-89]:

●In one report, 183 critically ill patients were divided into four groups of increasing severity according to the RIFLE classification scheme [88]. The most severe group, compared with all other groups, had significantly increased mortality at one and six months (74 and 86 percent, respectively). (See "Definition and staging criteria of acute kidney injury in adults".)

●Among 843 cardiac surgery patients, death was much more common in the 145 patients who developed postoperative AKI (14.5 versus 1.1 percent) [59]. Multivariable analysis revealed that in-hospital mortality was significantly associated with acute kidney failure (HR 7.8, 95% CI 3.1-20.0).

The impact of AKI on in-hospital mortality may be proportional to the degree of elevation in the serum creatinine independent of dialysis [90-93]. This issue was addressed in a single-center survey of 9210 hospitalized adults who had two or more serum creatinine concentrations measured [90]. An increase in the serum creatinine concentration of ≥0.5, ≥1, and ≥2 mg/dL (44, 88, and 177 micromol/L) was associated with an incrementally increased risk of in-hospital mortality (adjusted OR 6.5, 11, and 50, respectively) [90]. In addition, even milder forms of AKI adversely affect long-term mortality. In a large, retrospective study of 864,933 United States veteran patients with a mean follow-up time of 2.3 years, hospitalized patients surviving an episode of AKI for >90 days after discharge from the hospital and not requiring dialysis experienced an adjusted long-term mortality risk of 1.41 (95% CI, 1.39-1.43) [94].

The possibility that nonoliguric AKI is associated with a more favorable prognosis relative to oliguric AKI is discussed separately. (See "Nonoliguric versus oliguric acute kidney injury".)

CARDIOVASCULAR OUTCOMES — Severe AKI has been associated with an increase in cardiovascular events after discharge [95,96]. In an analysis of data provided by inpatient claims from Taiwan National Health Insurance, 4869 patients who recovered from dialysis-requiring AKI were matched with patients who did not develop AKI [95]. Compared with matched controls, the incidence of cardiovascular events (defined as a composite of nonfatal myocardial infarction [MI], coronary artery bypass graft, and coronary angiography) was higher among patients who developed AKI (hazard ratio [HR] 1.67, 95% CI 1.36-2.04). The increased cardiovascular risk appeared to be at least partly independent of subsequent progression of chronic kidney disease (CKD; although the strength of the association was stronger when progression to CKD was considered) [96]. It is possible that the increased cardiovascular risk is mediated by an increase in hypertension. (See 'Hypertension' below.)

Less severe AKI may not be associated with increased risk of cardiovascular events. In a study that compared patients admitted with AKI or AKI plus MI with those admitted with MI alone, those admitted with AKI alone had a lower risk of a subsequent major cardiovascular event (including congestive heart failure [HF], stroke, or MI) compared with those admitted with MI alone, with a HR of 0.40 (95% CI 0.38-0.44) [68]. In this study, AKI was defined either by the admitting International Classification of Disease (ICD)-9 code or by a minimum serum creatinine increase of 0.3 mg/dL developing within 48 hours or by a 50 percent increase developing over seven days, as defined by Kidney Disease: Improving Global Outcomes (KDIGO) [37]. As a result of this definition, some patients may have had less severe and transient AKI compared with the study from Taiwan that is cited above.

The association between cardiovascular risk and CKD is discussed elsewhere. (See "Chronic kidney disease and coronary heart disease".)

HYPERTENSION — Individuals who recover from AKI may be at an increased risk for hypertension. This was suggested by a retrospective cohort study of 2451 patients who had developed AKI during hospitalization [97]. After multiple adjustments for demographic factors, body mass index (BMI), preadmission blood pressure, smoking status, diabetes, chronic heart failure (HF), coronary heart disease, preadmission estimated glomerular filtration rate (eGFR), and proteinuria, AKI was associated with a 22 percent increase in risk of having blood pressure >140/90 mmHg measured in an ambulatory, nonemergency setting (95% CI 12-33 percent) as compared with control subjects who did not develop AKI.

SUMMARY

●Patients with acute kidney injury (AKI) have a maintenance phase that typically lasts between 7 and 21 days, although the duration is variable. The duration is dependent upon the length and severity of the initial ischemic episode; whether or not recurrent ischemia occurs or exposure to nephrotoxins is ongoing; and, perhaps, whether the patient is oliguric or nonoliguric. Whereas some patients recover within days, others require kidney replacement therapy (KRT) for weeks to months. (See 'Epidemiology' above and 'Duration of acute kidney injury' above.)

●Patients who recover from AKI may not return to their baseline kidney function. An irreversible decline in kidney function after recovery is more likely in patients over age 65 years, those with pre-existing chronic kidney disease (CKD), and those with heart failure (HF). Patients with CKD who develop AKI are more likely to progress to end-stage kidney disease (ESKD) compared with patients with CKD who do not experience an episode of AKI. (See 'Degree of recovery' above.)

●AKI during hospitalization is associated with high in-hospital and long-term mortality. A variety of factors have been associated with increased mortality, including male sex, race, older age, oliguria, sepsis, respiratory or liver failure, cerebrovascular events, and, in particular, overall severity of illness. (See 'Patient survival' above.)

●Patients who develop AKI in the hospital and are discharged alive should be evaluated within three months for resolution or new onset or worsening of pre-existing CKD. (See 'Degree of recovery' above.)