Acute hemodialysis prescription

INTRODUCTION — Acute kidney injury (AKI) is a major cause of morbidity and mortality, particularly in the hospital setting. Despite improvements in kidney replacement therapy (KRT) during the last several decades, the mortality rate associated with AKI in critically ill patients remains high. (See "Kidney and patient outcomes after acute kidney injury in adults".)

Acute KRT is commonly indicated for patients with AKI. Available modalities for acute KRT include peritoneal dialysis, intermittent hemodialysis and variations of intermittent hemodialysis (such as hemofiltration and prolonged intermittent kidney replacement therapy [PIKRT]), and continuous kidney replacement therapy (CKRT).

This topic reviews the acute hemodialysis prescription for patients with AKI. The indications for acute dialysis are discussed elsewhere.

●(See "Kidney replacement therapy (dialysis) in acute kidney injury in adults: Indications, timing, and dialysis dose", section on 'Urgent indications'.)

Other dialysis modalities are discussed separately.

●(See "Kidney replacement therapy (dialysis) in acute kidney injury in adults: Indications, timing, and dialysis dose", section on 'Optimal modality'.)

●(See "Continuous kidney replacement therapy in acute kidney injury".)

●(See "Use of peritoneal dialysis (PD) for the treatment of acute kidney injury (AKI) in adults".)

●(See "Prolonged intermittent kidney replacement therapy".)

COMPONENTS OF THE ACUTE HEMODIALYSIS PRESCRIPTION — The components of the acute dialysis prescription include the choice of hemodialysis membrane, dialysis session length, blood flow rate, dialysate composition and temperature, amount and rate of ultrafiltration (UF), and choice of anticoagulation.

Individual components of the prescription vary depending upon the indications for dialysis and on patient-specific variables.

Choice of dialyzer — There are three types of membranes used to manufacture dialyzers: cellulose, substituted cellulose, and noncellulose synthetic materials. The vast majority of dialyzers contain synthetic membranes. We use dialyzers with noncellulose synthetic membranes for acute hemodialysis. Cellulose membranes and, to a lesser degree, substituted cellulose membranes, are relatively bioincompatible; such membranes tend to initiate an inflammatory cascade and activate complement, resulting in the generation of the anaphylatoxins C3a and C5a. Synthetic membranes are more biocompatible and may be associated with fewer infectious complications and improved rates of recovery of kidney function compared with their cellulose counterparts. (See "Dialysis-related factors that may influence recovery of kidney function in acute kidney injury (acute renal failure)", section on 'Characteristics of the dialysis membrane'.)

There are a variety of noncellulose synthetic membranes available, including polyacrylonitrile (PAN), polysulfone, polycarbonate, polyamide, and polymethylmethacrylate. Except for PAN membranes, all are effective and safe. PAN membranes can activate the kallikrein-kinin cascade, leading to the generation of bradykinin, which is a mediator of pulmonary vasoconstriction and systemic hypotension [1]. We do not use PAN membranes (which are not available in the United States) for acute intermittent hemodialysis, although this membrane remains in use in some continuous kidney replacement systems outside of the United States.

If patients are being dialyzed with a PAN membrane, then the concomitant use of angiotensin-converting enzyme (ACE) inhibitors should be avoided since this combination has been associated with anaphylactoid reactions [2,3]. Such reactions may be related to a further potentiation of bradykinin response to PAN membranes in the presence of ACE inhibitors (which are also kinase inhibitors) [4,5]. By contrast, angiotensin receptor blockers are not kinase inhibitors and have not been associated with anaphylaxis with PAN membranes. ACE inhibitor-associated anaphylactoid reactions are far less common with surface-treated PAN/AN69 membranes [6].

The synthetic membranes listed above are high-flux membranes. High-flux membranes have greater permeability for larger molecules [7]. Clinical studies have not generally demonstrated a clear benefit from using high-flux membranes. (See "Dialysis-related factors that may influence recovery of kidney function in acute kidney injury (acute renal failure)", section on 'Characteristics of the dialysis membrane'.)

Low-flux cellulose dialyzers are either conventional (ie, low efficiency) or high efficiency. If conventional-efficiency, low-flux cellulose dialyzers are used, the dialysis time may need to be increased in order to achieve adequate clearances. (See "Kidney replacement therapy (dialysis) in acute kidney injury in adults: Indications, timing, and dialysis dose", section on 'Intermittent hemodialysis'.)

If a high-efficiency cellulose dialyzer is used, the urea clearance is similar to high-flux membranes, provided that blood flow is adequate. (See 'Blood flow rate' below and 'Dialysis dose' below.)

Dialysis session length — The duration of each hemodialysis session affects clearance of uremic toxins and correction of electrolyte abnormalities. A dialysis session of less than three hours may not replace enough alkali or remove enough potassium, phosphate, or uremic solutes. Acute hemodialysis treatments are most commonly three and a half to four hours, but the duration should be adjusted as clinically indicated. The dialysate composition and ultrafiltration goals should be modified, taking session length, frequency, and intensity into account. (See 'Dialysate composition' below and 'Ultrafiltration' below and 'Dialysis dose' below.)

Blood flow rate — For the first dialysis session, the blood flow rate should be based upon the elevation in blood urea nitrogen (BUN) prior to starting dialysis. There are no published data to guide an optimal approach.

●If the BUN is >100 mg/dL, we lower the blood flow rate and shorten the dialysis treatment time to avoid a rapid reduction of BUN and plasma osmolarity, which can lead to dialysis dysequilibrium syndrome (DDS). The blood flow and treatment time are then gradually increased over several consecutive days to the maximum of 400 mL/min. This approach to preventing DDS is discussed in more detail elsewhere. (See "Dialysis disequilibrium syndrome", section on 'Patients being newly initiated on hemodialysis'.)

●If aggressive dialysis is needed for other reasons, such as severe hyperkalemia, we are not as conservative in terms of modifying blood flow and treatment time. Additional details about dialysis dosing are presented separately. (See "Kidney replacement therapy (dialysis) in acute kidney injury in adults: Indications, timing, and dialysis dose", section on 'Optimal dosing'.)

In addition, for patients with severe azotemia and those with, or at risk for, increased intracranial pressure, continuous kidney replacement therapies (CKRTs) may be used; compared with acute intermittent hemodialysis, lower blood flows are sufficient to achieve adequate clearance using CKRT due to its continuous nature [7]. (See "Continuous kidney replacement therapy in acute kidney injury".)

Dialysate composition — Dialysate solutes include potassium, sodium, bicarbonate (or other buffer), calcium, magnesium, chloride, and glucose. Unlike chronic hemodialysis (in which the dialysate composition is generally the same for every treatment), the dialysate composition in acute hemodialysis may need to be altered for each treatment to correct the metabolic abnormalities that can rapidly develop during acute kidney injury (AKI). This is particularly common as marked potassium and/or acid-base derangements that may have led to dialysis initiation are controlled.

Potassium — Patients with severe hyperkalemia are often treated medically prior to dialysis. This issue is discussed separately. (See "Treatment and prevention of hyperkalemia in adults", section on 'Patients with a hyperkalemic emergency'.)

There is no high-quality evidence to inform the selection of dialysate potassium concentration. Recommendations are based upon understanding of the physical principles underlying dialysis, clinical observations, and on the limited studies performed in maintenance dialysis population [8-11].

The typical potassium concentration in the dialysate for acute hemodialysis ranges from 2 to 4 mEq/L. The prescribed dialysate potassium concentration is based upon the predialysis serum potassium value, other clinical matters such as ongoing tissue ischemia and cell lysis, session length, and frequency [12].

The rate of correction of serum potassium will depend upon the serum to dialysate potassium concentration gradient and the aggressiveness of other components of the dialysis prescription, including flow rates, dialyzer surface area, and alkali concentration. (See 'Laboratory testing' below and 'Efficiency of potassium removal' below.)

Choice of dialysate potassium concentration — Although there is no general consensus concerning the optimal strategy, the following is our approach to adjusting the dialysate potassium concentration. The serum potassium concentrations noted below are arbitrary, and other approaches are likely to be equally effective and safe.

●Serum potassium <4.5 mEq/L – If the predialysis serum potassium level is <4.5 mEq/L, we use a dialysate potassium concentration of 3 to 4 mEq/L. This concentration prevents the development of hypokalemia. Some clinicians prefer a slightly lower net dialysate potassium concentration if the serum potassium is 4 to 4.5 mEq/L. Patients with a predialysis serum potassium level <3.5 mEq/L may require additional potassium supplementation and should have a serum potassium level checked prior to each dialysis session.

●Serum potassium between 4.5 and 5.5 mEq/L – If the predialysis serum potassium level is between 4.5 and 5.5 mEq/L, we use a dialysate potassium of 3 mEq/L.

However, if the patient has an ongoing reason for hyperkalemia (eg, marked rhabdomyolysis), then we may use a lower dialysate potassium of 2 mEq/L to decrease the risk of the patient developing hyperkalemia prior to the next dialysis session. Among such patients, the serum potassium may be low to normal following dialysis, however, and we usually do not do this among patients who are at risk for arrhythmias related to potassium removal [13-16]. For patients at increased risk of arrhythmias (eg, those with coronary artery disease, left ventricular hypertrophy, digoxin use, hypertension, or advanced age), some nephrologists even avoid using a dialysate potassium <3 mEq/L. However, the increased safety of a 3 compared with a 2 mEq/L potassium dialysate in this situation has not been shown. Patients who have ongoing risks of hyperkalemia and are at risk for arrhythmias may benefit from CKRT rather than intermittent hemodialysis. (See 'Telemetry' below.)

●Serum potassium between 5.5 and 8 mEq/L – For most patients with a predialysis potassium level between 5.5 and 8 mEq/L, we use a 2 mEq/L dialysate potassium bath. As noted above, some nephrologists do not use a dialysate potassium <3 mEq/L for patients who are at risk for arrhythmias.

●Serum potassium >8 mEq/L – For patients with severe hyperkalemia (eg, >8 mEq/L), we use a dialysate potassium concentration of 2 mEq/L in order to rapidly decrease the serum potassium to a safer level. However, for such patients, some nephrologists use a 1 mEq/L potassium dialysate, which is effective but does require close monitoring.

The advantage of using a 2 mEq/L over a 1 mEq/L dialysate potassium concentration is the lower risk of overly rapid shifting of potassium, thereby reducing the risk of myocardial instability. In addition, a higher serum-dialysate potassium gradient can result in a more rapid postdialysis rebound of potassium.

All patients who are being dialyzed with a dialysate potassium concentration of 1 mEq/L should be monitored on telemetry for arrhythmias and have the serum potassium measured every 30 to 60 minutes during dialysis. Once the serum potassium is between 6 and 7 mEq/L, the dialysate potassium concentration can be changed to 2 mEq/L for the remainder of the hemodialysis session; this is less likely to cause hypokalemia and avoids a large blood-dialysate potassium difference, which may contribute to the risk of arrhythmia [17].

We do not use a zero potassium bath, even for severe hyperkalemia, although it is most effective in reducing the serum potassium in a short period of time [18,19]. A zero potassium bath has been associated with an increased risk of hypokalemia and dialysis-induced arrhythmias although rigorous data are lacking [17,19].

Efficiency of potassium removal — The rate of potassium removal by dialysis is affected by multiple variables.

●Potassium rebound – Potassium is transferred from serum into the dialysate during dialysis. Serum potassium decreases quickly during the initial one to two hours of the treatment. Later during the dialysis session and afterward, there is flux of potassium from the intracellular space to the serum. This can lead to a marked rebound in the serum potassium concentration within one to two hours after dialysis.

●Effect of rapidly acting therapies for hyperkalemia – The amount of potassium removal is equal to the amount removed by ultrafiltration (convection) and the amount removed by diffusion. The rate of diffusion is proportional to the gradient between the serum and dialysate concentrations. The administration of insulin, intravenous (IV) glucose, and beta agonists, either concurrently or prior to hemodialysis, all lower the rate of potassium removal during dialysis because they cause intracellular translocation of potassium, which lowers the serum potassium concentration. Such therapies, if used, should be stopped shortly before dialysis is initiated. The approach to hyperkalemia prior to dialysis is discussed elsewhere. (See "Treatment and prevention of hyperkalemia in adults".)

●Effect of dialysate glucose – The dialysate glucose concentration may modulate potassium removal since the glucose load enhances insulin secretion, which drives potassium into the cells. Dialysate glucose concentrations of either 100 or 200 mg/dL are considered "standard." In one study, dialysis against a dialysate glucose concentration (ie, 200 mg/dL [11.1 mmol/L]) resulted in less potassium removal compared with glucose-free dialysate solution [20]. Thus, in cases of severe hyperkalemia where potassium removal is critical, a lower dialysate glucose concentration (ie, 100 mg/dL [5.5 mmol/L]) may theoretically be used to facilitate potassium removal. However, this is usually not done in practice. (See 'Glucose' below.)

●Effect of metabolic acidosis – Metabolic acidosis also affects serum potassium levels. The correction of metabolic acidosis allows extracellular potassium to enter cells and may speed the correction of hyperkalemia without net potassium removal.

Sodium — The choice of dialysate sodium concentration for individual patients depends upon the predialysis serum sodium concentration and the hemodynamic status of the patient.

The dialysate sodium may affect hemodynamic stability during acute hemodialysis. This is discussed below. (See 'UF-related hypotension' below.)

Our approach is based upon the predialysis serum sodium concentration.

Normal serum sodium — For patients with normal or near-normal serum sodium levels, we use a dialysate sodium concentration of 137 to 138 mEq/L.

The dialysate sodium concentration that results in no net diffusive transfer of sodium is generally 0.1 to 3 mEq/L below that of the predialysis serum sodium concentration [21-24] although there may be significant differences between prescribed and measured bath sodium concentrations [25].

Chronic hyponatremia — Severe chronic hyponatremia may be safely corrected with hemodialysis; however, rapid correction must be avoided [26]. Rapid correction of severe chronic hyponatremia can lead to osmotic demyelination (pontine and extrapontine myelinolysis). Uremia may provide some protection against osmotic demyelination. However, there is a case report of osmotic demyelination following dialysis of a severely hyponatremic patient [27].

Another approach that can help mitigate the risk of rapid correction of hyponatremia in cases that are not urgent is to medically treat for the first one to two days. This can result in raising the serum sodium concentration, thereby reducing the serum-dialysate sodium gradient.

Definitions of chronicity and the recommended rate of correction are the same as those in the nondialysis general population. (See "Overview of the treatment of hyponatremia in adults".)

Among patients with severe chronic hyponatremia (ie, <120 mEq/L), we adjust the dialysate sodium to prevent rapid correction [28]. We set the dialysate sodium to the lowest commercially available setting (130 mEq/L), reduce the blood flow rate to 2 mL/kg/min, and reduce the hemodialysis session length [29].

Because of the low blood flow rate, the sodium concentration of the blood returning to the patient rises to equal the sodium concentration of the dialysate (130 mEq/L). The expected rate of rise of the serum sodium concentration can be estimated from the rate of sodium infusion into the patient (which is the product of the concentration gradient between blood and dialysate and the blood flow rate) divided by total body water. However, hourly measurements of the serum sodium concentration during the course of dialysis are mandatory, and administration of small amounts of 5 percent dextrose in water (D5W) may still be required to assure that correction does not exceed 6 mEq/L during the dialysis treatment.

Change in Na (mEq/L)  =  [(Blood flow rate  x  Time on hemodialysis in minutes)  x  (Dialysate Na – Plasma Na)]  ÷  Total body water

The goal is to correct the hyponatremia over the course of multiple hemodialysis sessions that are performed over a period of several days.

CKRT may also be used to safely correct hyponatremia. CKRT is less efficient in the rate at which serum sodium is changed and results in a more gradual correction over a longer time span [30-33]. (See "Continuous kidney replacement therapy in acute kidney injury".)

Chronic hypernatremia — Among patients requiring acute dialysis, our approach to chronic hypernatremia depends upon its severity.

●If the serum sodium concentration is only mildly elevated, we use a dialysate sodium concentration that is within 2 mEq/L of the serum sodium concentration for the first dialysis session. The use of dialysate sodium concentrations more than 3 to 5 mEq/L below the serum sodium concentration is associated with hypotension, muscle cramps, and, most importantly, disequilibrium syndrome. Subsequently, correction of the hypernatremia is performed with the administration of hypotonic solutions. (See "Treatment of hypernatremia in adults".)

●Patients with extremely high serum sodium concentrations are best treated with CKRT [34,35]. Rapid correction of severe chronic hypernatremia should be avoided as overcorrection may lead to cerebral edema, although one study did not find this to be the case [26,36]. (See "Continuous kidney replacement therapy in acute kidney injury".)

Acute hypo- or hypernatremia — Patients with hyperacute salt poisoning (eg, due to the suicidal ingestion of sodium chloride or the inadvertent IV infusion of hypertonic saline during a therapeutic abortion) or hyperacute water intoxication (eg, as a complication of marathon running or use of the drug "Ecstasy") should undergo aggressive correction of their serum sodium concentration. Rapid correction is well tolerated in hyperacute disturbances. Conventional hemodialysis with a standard sodium concentration can be used to correct the electrolyte disturbance rapidly. (See "Overview of the treatment of hyponatremia in adults".)

Bicarbonate — The main dialysate buffer used in intermittent hemodialysis is bicarbonate. Bicarbonate is inexpensive and generally well tolerated. Although acetate used to be the predominant buffer used in hemodialysis, it is no longer routinely used, because it is associated with cardiac and hemodynamic instability.

A disadvantage of bicarbonate is that it precipitates as an insoluble salt when stored with the divalent cations calcium and magnesium [37]. As a result, the buffer and electrolytes are stored separately prior to hemodialysis [37]. Possible side effects of bicarbonate are related to the increase in pH induced by dialysis and include hypoxemia due to decreased respiratory drive and altered mental status, weakness, cramping, and lethargy [38].

Choosing the appropriate dialysate bicarbonate concentration is based upon the acid-base status of the patient. The acid-base status should be assessed using both the serum bicarbonate and pH. (See "Simple and mixed acid-base disorders".)

Our approach is based upon our clinical experience, and we stress that there are no published data to support these recommendations.

Patients with metabolic acidosis — Our approach to metabolic acidosis depends upon its severity and cause:

●For patients who have mild or moderate metabolic acidosis (ie, serum bicarbonate 10 to 23 mEq/L and an acidemic pH) or who have no acid-base disorder, we generally use a standard dialysate bicarbonate concentration of approximately 30 to 35 mEq/L.

●For patients with severe metabolic acidosis (ie, serum bicarbonate <10 mEq/L and a severely acidemic pH [eg, <7.2]), we use a dialysate bicarbonate solution of approximately 35 to 40 mEq/L. For such patients, an extended duration of hemodialysis may be necessary.

The use of this high dialysate bicarbonate concentration may result in the slower removal of potassium due to its intracellular translocation [39]. In addition, among patients who are treated with opioids, use of a high dialysate bicarbonate may increase opioid distribution into the central nervous system.

Patients with alkalosis — The severity of the alkalemia and the process generating the alkalosis are the main determinants of the dialysate bicarbonate concentration. In particular, the clinician should investigate whether there is ongoing generation versus a one-time insult causing the alkalosis. A one-time insult can be resolved with a single hemodialysis treatment whereas ongoing generation of alkalosis may require frequent and/or long hemodialysis sessions with a lower bicarbonate dialysate.

Both the blood pH and serum bicarbonate should be determined to appropriately assess the degree of alkalosis.

●If the predialysis serum bicarbonate level is >28 mEq/L or respiratory alkalosis is present, we use a bicarbonate concentration of 25 to 30 mEq/L [12]. We do not use a dialysate bicarbonate lower than 25 mEq/L.

●If this dialysate does not address ongoing alkalosis, we administer 1 to 3 liters of 0.9 percent sodium chloride (NaCl) per dialysis session while removing volume as necessary. Infusion of larger volumes of 0.9 percent NaCl may require adjustments to the dialysate sodium concentration, depending on daily electrolyte measurements and clinical evaluation. The administration of chloride often corrects the alkalosis although the entire mechanism by which this occurs is not clear. One mechanism is that the ultrafiltration of fluid containing a higher concentration of bicarbonate is replaced by infusion of a bicarbonate-free solution (ie, NaCl).

Patients on mechanical ventilation — Among patients who are unable to increase their ventilation, the administered bicarbonate with dialysis may result in higher pCO₂ values. The pCO₂ of the dialysate with bicarbonate dialysate can be as high as 100 Torr. If the patient's pulmonary function is intact, hyperventilation provoked by the increased pCO₂ may decrease hypoxia. However, if the patient is on ventilatory support, the ventilator settings may need to be adjusted to correct hypercapnia. In a patient that is not completely dependent on the ventilator, the hypercapnia may serve as a stimulus for improved ventilation and facilitate weaning and extubation.

In addition, in patients being mechanically ventilated using low-tidal volume ventilation, an increased dialysate bicarbonate concentration may be required to increase the serum bicarbonate concentration to compensate for the respiratory acidosis resulting from "permissive hypercapnia." In such patients with permissive hypercapnia, the goal should not be normalization of pH; we arbitrarily aim for a target pH of ≥7.2 at the end of a dialysis treatment. By contrast, in patients being mechanically hyperventilated to compensate for metabolic acidosis, the minute ventilation (respiratory rate and/or tidal volume) may need to be reduced to avoid severe alkalemia as the metabolic acidosis is corrected with dialysis.

Calcium — The dialysate calcium ranges from 2 to 3.5 mEq/L and is adjusted based on the serum calcium. Since total plasma calcium levels are poorly predictive of the ionized level (which is the clinically active value), particularly in critically ill patients and those with hypoalbuminemia, the plasma ionized calcium level should be measured prior to hemodialysis in patients with significant hypocalcemia or hypercalcemia. If the ionized calcium cannot be obtained, the measured total plasma calcium level should be corrected based upon the serum albumin level since the total plasma calcium concentration will change in parallel to the albumin concentration; this correction, however, will not account for changes in acid-base status. (See "Relation between total and ionized serum calcium concentrations".)

Our approach is based upon clinical experience:

●For patients with mild hypocalcemia, normocalcemia, or mild hypercalcemia (total plasma calcium level between 8 to 12 mg/dL [2 to 3 mmol/L], corrected for hypoalbuminemia]), we use a dialysate calcium concentration of 2.5 mEq/L.

●For patients with significant hypocalcemia (total plasma calcium level <8 mg/dL [<2 mmol/L], corrected for hypoalbuminemia), particularly if the patient is symptomatic, we use a dialysate calcium concentration of 3 to 3.5 mEq/L.

●For patients with severe hypercalcemia (total plasma calcium level >12 mg/dL [>3 mmol/L], corrected for hypoalbuminemia), we use a dialysate calcium concentration of 2 to 2.5 mEq/L.

There are limited data to inform the selection of dialysate calcium. Most of the literature is in the end-stage kidney disease (ESKD) setting and concerns the role of calcium in bone/mineral metabolism or in cardiovascular disease. In patients with AKI, or hospitalized patients with ESKD in whom acute dialysis is required, immediate cardiac concerns predominate, and the outcomes data derived from outpatients are less pertinent. The major concern in acute hemodialysis is that lower bath calcium concentrations may prolong and increase the variability of the QTc interval, both risk factors for sudden death [16]. The serum calcium can also influence potassium-induced arrhythmias [10].

Magnesium — The usual dialysate magnesium concentration is 0.5 to 1 mEq/L. Either concentration will address hypermagnesemia. Hypomagnesemia is usually corrected with IV or oral supplementation. (See "Hypomagnesemia: Evaluation and treatment".)

Glucose — The standard dialysate glucose concentration is 100 to 200 mg/dL (5.5 to 11.1 mmol/L). The relative benefits of using a lower as compared with a higher glucose concentration have not been evaluated in the acute setting.

Dialysate temperature — The temperature of the dialysate is typically set between 35 and 37 degrees Celsius. The lower range may be preferred in patients at risk for hypotension during dialysis or who have had hypotension during dialysis. Cooler dialysate temperatures, however, may be uncomfortable for some patients. There is evidence that chronic dialysis with cooler dialysis fluid may slow white matter changes in the brain [40].

Ultrafiltration

Determining UF goal — Determining optimal ultrafiltration (UF) requirements in critically ill patients with AKI is challenging. Volume status and the desired UF requirement are determined in part by physical examination and hemodynamic indices. In general, no one specific test or parameter is sufficient in isolation. As previously mentioned, tolerance to the rate of UF is one of the most important aspects of the hemodialysis prescription. (See 'Dialysis session length' above.)

The following two overriding principles should be recognized:

●The approach to volume overload is different for patients on chronic dialysis than for critically ill patients with AKI. The target weight of a patient on chronic dialysis is usually determined empirically as the weight at which clinical signs of extracellular fluid expansion are absent and below which clinical signs of extracellular fluid depletion arise. By contrast, among critically ill patients with AKI, the volume expansion that is frequently observed is often necessary to maintain optimal circulatory and oxygen transport status.

●The relationship between blood volume and hypotension is different in patients on chronic dialysis and critically ill patients with AKI. Among critically ill patients with AKI, the relationship between volume status and hemodynamic stability is unpredictable. As an example, blood volume monitoring, a biofeedback system that automatically adjusts UF rate and dialysate sodium content in response to a fall in circulating intravascular volume, reduces the occurrence of intradialytic hypotension in patients with ESKD [41] but is ineffective for preventing hypotension in critically ill patients with AKI [42]. This lack of a predictable relationship between volume status and hemodynamic stability means that UF goals for a given patient should be assessed not only in terms of fluid mass balance or the mandatory removal of obligatory fluid loads but also in terms of the effect of intervention on the patient's broader clinical condition and hemodynamic status.

In hemodynamically stable patients, the estimation of target intravascular volume can be made in the usual fashion utilized for patients on chronic dialysis. However, in hemodynamically unstable patients, some experts use invasive or noninvasive monitoring (such as bioimpedance analysis, pulse contour analysis, echocardiography, or inferior vena cava measurement by ultrasound) to guide the UF goals for a given intermittent hemodialysis session.

UF-related hypotension — Ultrafiltration (UF) during hemodialysis can result in significant intradialytic hypotension. This can be treated by reducing or discontinuing UF. (See "Intradialytic hypotension in an otherwise stable patient".)

Additional measures that may help prevent intradialytic hypotension during acute hemodialysis in AKI include the following:

●Increasing the frequency and/or duration of treatments – Increasing the frequency of treatments may allow a decrease in UF per session; however, this may not result in a decreased risk of intradialytic hypotension. In one study that randomly assigned patients to intermittent hemodialysis either three or six times per week, the percentage of dialysis sessions complicated by intradialytic hypotension and the severity of intradialytic hypotension were similar in both treatment arms, with more patients sustaining intradialytic hypotension with the more frequent treatment schedule [43]. Some trials comparing intermittent hemodialysis with CKRT for AKI have used longer dialysis treatments (eg, five to five and a half hours) to minimize hypotension in hemodynamically unstable patients [44]. Changes to the frequency and/or duration of treatments may depend upon nursing availability and other logistic considerations.

●Sodium and UF modeling – Sodium modeling is a method by which a higher dialysate sodium concentration is used at the beginning of hemodialysis and progressively decreased throughout the session to avoid abruptly lowering the plasma osmolarity [38]. UF modeling is an automated method by which the set UF rate is higher early in the hemodialysis procedure and lower later in the procedure or some modification of this [45]. Combination sodium and UF modeling has been studied but has not been conclusively shown to be beneficial [46]. It is usually used because of its ease of application. (See "Intradialytic hypotension in an otherwise stable patient", section on 'Prevention of recurrent episodes'.)

We prefer either of the following two specific strategies:

•With one high/low-sodium modeling prescription, a high-dialysate sodium (eg, 150 mEq/L) alternates with a low-dialysate sodium (eg, 130 mEq/L), with each level set for an equal amount of time. The average of the high/low-sodium levels (eg, 140 mEq/L) is the dialysate sodium usually prescribed in hemodynamically stable patients with normal serum sodium levels. During the low-sodium period, the UF rate is minimized or stopped. UF only occurs during the high-sodium period to draw out intracellular water due to the extracellular hypernatremia. This program is widely available in Europe but not in some machines in the US.

•Another sodium modeling prescription is to set the initial dialysate sodium at a high level (eg, 150 to 160 mEq/L). The dialysate sodium level is then decreased in stepwise, exponential, or linear decrements (depending on clinical effect) to a final low level (eg, 140 mEq/L). To maintain isonatremia, the time-average concentration of dialysate sodium should be the same or marginally lower than the predialysis serum sodium concentration (approximately within 1 to 2 mEq/L). With a linear sodium profile, for example, the duration (and degree) of dialysis spent below the isonatremic concentration must be approximately equal to that spent above it [21].

Anecdotal experience has suggested that, in addition to the above approaches, using a slightly higher sodium of 143 mEq/L or so throughout the entire dialysis rather than 137 to 138 mEq/L may be effective in increasing hemodynamic stability. This will raise the plasma sodium concentration, at least temporarily.

Although sodium modeling may be associated with improved hemodynamic stability acutely, if dialytic support is prolonged, sodium modeling may be associated with a positive sodium balance due to diffusion of sodium from dialysate to blood because of the higher sodium concentration in the dialysate used during sodium modelling. This may lead to difficulty in blood pressure and edema management.

The effects of sodium and UF modeling on hemodynamic stability during acute hemodialysis were demonstrated in a randomized, crossover trial of 10 patients with AKI in the intensive care unit (ICU) [47]. The study used either a fixed dialysate sodium regimen (140 mEq/L) and fixed UF rate or a variable sodium dialysate (160 to 140 mEq/L) and variable UF rate (such that half of the fluid was removed during the first third of the treatment and the remaining half over the last two-thirds). Compared with patients on the fixed regimen, those on sodium and UF modeling had greater hemodynamic stability and fewer interventions involving nursing and volume replacement.

Multiple sodium modeling prescriptions are programmed in most hemodialysis machines though the availability of specific programs may differ in machines used in Europe versus the United States. Patients may respond to only one or all available prescriptions. Thus, trials are required to find the best sodium modeling prescription in patients with AKI on hemodialysis.

It is important to note that a lower dialysate sodium may facilitate sodium efflux and cause less interdialytic weight gain. It can, however, introduce more intradialytic hypotension and patient-reported symptoms. On the other hand, a higher dialysate sodium may result in "sodium loading" the patient, with subsequent increased patient thirst, weight gain, and blood pressure. Most studies pertaining to sodium modeling were in patients on chronic dialysis; they may be applicable to patients who are admitted to the inpatient setting for a prolonged period of time.

●Cool-temperature dialysis – Cool-temperature dialysis may increase hemodynamic stability. Data in support of this intervention are largely derived from the chronic hemodialysis setting. The method for performing cool-temperature dialysis is the same as that among chronic dialysis patients and is presented separately. (See "Intradialytic hypotension in an otherwise stable patient", section on 'Second-line approach'.)

However, cool-temperature dialysis can cause patients to feel chilled and shiver, and hypothermia may have adverse effects on myocardial function, end-organ perfusion, blood clotting, and, possibly, kidney function recovery [48].

●Higher dialysate calcium concentration – Increasing the dialysate calcium may be used in combination with some of the aforementioned interventions to treat intradialytic hypotension. However, we do not use a dialysate calcium concentration >3.5 mEq/L. (See 'Calcium' above.)

In a prospective, crossover study that included patients who were also on midodrine, cool dialysate, or a combination of these two therapies, compared with low-dialysate calcium, the use of high-dialysate calcium increased post-hemodialysis mean arterial pressure but did not reduce symptoms or interventions for intradialytic hypotension [49].

●Other measures – Despite the measures listed above, hemodynamic instability may still occur because of the various dialysis-independent causes of intradialytic hypotension present in the acute setting (eg, cardiogenic, vasodilatory, or hypovolemic shock). Other measures that can be used include the following:

•Performing hemodialysis with vasopressor support may be an option if the pressor support keeps the patient hemodynamically stable during their treatment. Midodrine (alpha-1 adrenergic agonist used in autonomic dysfunction) may be administered in patients who do not require more powerful, pharmacologic forms of pressor support. (See "Intradialytic hypotension in an otherwise stable patient", section on 'Third-line approach'.)

•IV boluses of 0.9 percent NaCl or albumin [50] given during hemodialysis can transiently increase blood pressure.

•If measures to improve hemodynamic stability during intermittent hemodialysis sessions are not successful, switching to prolonged intermittent kidney replacement therapy (PIKRT) or CKRT usually improves hemodynamics while maintaining an acceptable rate of UF and solute clearance. (See "Prolonged intermittent kidney replacement therapy" and "Continuous kidney replacement therapy in acute kidney injury".)

Anticoagulation — Issues surrounding anticoagulation in patients undergoing acute hemodialysis are presented separately. (See "Anticoagulation for the hemodialysis procedure".)

Dialysis dose — Dialysis dose in AKI is discussed elsewhere. (See "Kidney replacement therapy (dialysis) in acute kidney injury in adults: Indications, timing, and dialysis dose", section on 'Intermittent hemodialysis'.)

MONITORING DURING ACUTE HEMODIALYSIS

Telemetry — Patients with underlying cardiac disorders who undergo acute hemodialysis should be placed on a cardiac rhythm monitor during the dialysis session. Dialysis-induced reductions in the serum potassium can provoke ventricular arrhythmias [13-17,51].

Other measures to reduce the risk of arrhythmias include close monitoring of the serum potassium and avoidance of very low dialysate potassium, as noted above. (See 'Choice of dialysate potassium concentration' above.)

Laboratory testing — A basic metabolic profile (electrolytes, glucose, urea, creatinine, phosphate, calcium, and magnesium) should be reviewed prior to acute hemodialysis sessions since electrolyte and acid/base status can profoundly change between treatments and require alterations in the dialysis prescription. In some patients, it may also be appropriate to obtain a basic metabolic profile or other blood chemistries sometime after the end of dialysis to monitor the response to the treatment. Postdialysis chemistry studies, if indicated, should generally be done after the first one to two hours following completion of the session to allow for equilibration between the plasma and intracellular compartments.

MANAGEMENT DURING RECOVERY OF KIDNEY FUNCTION — Most patients who survive acute kidney injury (AKI) recover some degree of their kidney function [52,53]. Patients should be monitored closely for recovery and dialysis discontinued when acceptable kidney function is restored.

Monitoring for recovery — A key element to successful discontinuation of dialysis is monitoring for recovery of kidney function. In the hospital, predialysis laboratory tests should include serum potassium, bicarbonate, creatinine, and blood urea nitrogen (BUN). Recovery of kidney function may be heralded by an increase in spontaneous or diuretic-induced urine volume and/or declining predialysis BUN and creatinine concentrations. For patients whose urine output is greater than 400 mL/day, some clinicians perform a timed urine collection (on an interdialytic day) or interdialytic urine collection for creatinine and urea clearance to guide decision making.

Discontinuation of dialysis — Continued hemodialysis may be detrimental to recovery of kidney function. Hypotension associated with dialysis may reduce kidney function [54,55], and morphologic studies in humans have shown fresh ischemic lesions in kidney biopsy specimens from patients with AKI of over three weeks duration [56].

There are no accepted standards for discontinuation of dialysis [57-60]. In some cases, recovery of kidney function is obvious, as indicated by rapidly increasing urine output and/or decreasing levels of predialysis serum creatinine and/or BUN values. In many cases, however, recovery is both slow and sporadic.

●For those patients with AKI who are otherwise medically recovered from their acute illness, kidney function alone is usually sufficient to guide dialysis requirements. Such patients rarely require dialysis when their glomerular filtration rate (GFR) is >10 to 15 mL/min, particularly when their predialysis serum creatinine and/or BUN values are decreasing, unless dialysis is needed for management of volume overload [43,61]. In this setting, we generally use the serum creatinine, rather than the BUN, because the creatinine is less confounded by the effects of metabolism.

●For patients who are acutely ill, the decision to discontinue dialysis should not be made solely on the presence or degree of recovery of kidney function [57,60]. The decision should consider the patient's overall condition (including presence of increased catabolism, fluid overload, ongoing hemodynamic instability, and ongoing requirement for nephrotoxic drugs or large volumes of fluid). It should be noted that the recovery of electrolyte homeostasis does not always parallel the recovery of GFR, so electrolyte abnormalities may also determine the need for dialytic support during the AKI recovery phase. As GFR improves, water homeostasis may not have been restored, and an osmotic diuresis may result in polyuria that must be managed.

Over half of patients with AKI will be able to successfully stop dialysis on their first attempt, and others will need reinitiation of treatment, albeit at lower doses [52,62,63]. Urine output and duration of the need for dialysis have been reported to be important predictors of successful discontinuation of dialysis. Stopping dialysis is less likely to be successful if urine output is less than 400 to 600 mL/day (without diuretics) [52,63].

There are no studies that suggest an optimal approach to withdrawing dialytic support. We generally tailor our approach to the individual patient and expected patient response. For those with clinically obvious and rapid recovery of kidney function, abrupt cessation of dialysis is appropriate and well tolerated, as long as adequate monitoring is in place.

However, for those who are still acutely ill or who have had a very slow or sporadic recovery, we generally wean dialysis by decreasing the frequency of treatments as tolerated, which is usually over several weeks. This approach may also be optimal for those with preexisting chronic kidney disease (CKD) or those with other important comorbidities such as congestive heart failure, where recovery of kidney function might be compromised.

Use of loop diuretics during recovery — Patients with high interdialytic weight gains (despite improved GFR) may benefit from a trial of loop diuretic therapy while the dialysis duration or frequency is being decreased.

The addition of diuretics does not enhance kidney function recovery or improve kidney function [64,65] but may increase urine output and sodium excretion [66-68] and allow for a lower ultrafiltration (UF) during a dialysis session [69].

Two clinical trials and one observational study have shown that the addition of high-dose diuretics in patients with AKI does not improve weaning from kidney replacement therapy (KRT) in the hospital or intensive care unit (ICU) setting [63,68,70]. Diuretic therapy should therefore not be routinely used for this purpose [58].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Dialysis".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Hemodialysis (The Basics)")

●Beyond the Basics topic (see "Patient education: Hemodialysis (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●General principles – Acute kidney injury (AKI) is a major cause of morbidity and mortality, particularly in the hospital setting. Acute kidney replacement therapy (KRT), including acute intermittent hemodialysis, is commonly indicated for patients with AKI. (See 'Introduction' above.)

●Acute hemodialysis prescription – The components of the acute dialysis prescription include the following:

•Choice of dialyzer – Biocompatible high-flux membranes are typically used for acute hemodialysis. However, the choice of dialyzer may depend upon what is available at the particular dialysis facility. (See 'Choice of dialyzer' above.)

•Dialysis session length – The duration of each hemodialysis session affects clearance of uremic toxins and correction of electrolyte abnormalities. Acute hemodialysis treatments are most commonly three and a half to four hours, but the duration should be adjusted as clinically indicated. (See 'Dialysis session length' above.)

•Blood flow rate – For the first dialysis treatment, the blood flow rate depends upon the degree of azotemia. If the blood urea nitrogen (BUN) is >100 mg/dL, the blood flow rate should be lowered and the dialysis treatment time shortened to prevent the development of dialysis dysequilibrium syndrome (DDS). This is discussed elsewhere. (See "Dialysis disequilibrium syndrome", section on 'Patients being newly initiated on hemodialysis'.)

If aggressive dialysis is needed for other reasons, such as severe hyperkalemia, we are not as conservative in terms of modifying blood flow and treatment time. (See 'Blood flow rate' above.)

•Dialysate composition and temperature – Dialysate components include potassium, sodium, bicarbonate buffer, calcium, magnesium, chloride, and glucose. The dialysate composition is routinely altered to correct the metabolic abnormalities that can rapidly develop during AKI. The temperature of the dialysate is typically set between 35 and 37 degrees Celsius. (See 'Dialysate composition' above and 'Dialysate temperature' above.)

•Ultrafiltration (UF) – The UF requirement is determined by physical examination and hemodynamic indices. In hemodynamically unstable patients, target intravascular volume should be titrated to invasive or noninvasive monitoring. UF may cause intradialytic hypotension, which can be treated by reducing or discontinuing UF, reducing the blood flow rate, and/or using additional measures. (See 'Determining UF goal' above and 'UF-related hypotension' above.)

•Anticoagulation and dialysis dose – These issues are discussed separately. (See "Anticoagulation for the hemodialysis procedure" and "Kidney replacement therapy (dialysis) in acute kidney injury in adults: Indications, timing, and dialysis dose", section on 'Intermittent hemodialysis'.)

●Management during recovery of AKI – Most patients who survive AKI recover kidney function. Dialysis should not be continued for longer than is necessary. For patients who are still acutely ill, the decision to discontinue dialysis should be determined by the degree of recovery of kidney function and by the patient's overall condition (including presence of increased catabolism, fluid overload, ongoing hemodynamic instability, and ongoing requirement for nephrotoxic drugs or large volumes of fluid). (See 'Management during recovery of kidney function' above.)

ACKNOWLEDGMENT — The editorial staff at UpToDate acknowledges Mark R Marshall, MD, and Phillip Ramos, MD, MSCI, who contributed to earlier versions of this topic review.

The editorial staff also acknowledges Gerald Schulman, MD, FASN, now deceased, who contributed to an earlier version of this topic.