INTRODUCTION — End-stage kidney disease (ESKD) requiring dialysis is a growing problem worldwide (see "Epidemiology of chronic kidney disease"). Dialysis-dependent patients commonly require surgery for reasons related to ESKD, including vascular access procedures, parathyroidectomy, or renal transplantation. Patients may also require elective or emergent surgical procedures for reasons unrelated to their ESKD.

This topic reviews the preanesthesia consultation and anesthetic management of patients on dialysis. Preoperative medical management of these patients is discussed separately. (See "Medical management of the dialysis patient undergoing surgery".)

PREANESTHESIA ASSESSMENT AND MANAGEMENT

Dialysis considerations — In patients with end-stage kidney disease (ESKD), renal replacement therapy is used to remove excess intravascular and extravascular volume, metabolic waste products, and normalize acid-base and electrolyte values. For those patients undergoing elective surgery who are on maintenance hemodialysis, a hemodialysis treatment is usually performed when practical, either the day before or on the day of the procedure. For those on peritoneal dialysis, their dialysis treatments are generally performed until just prior to the procedure. Peritoneal dialysate should be drained prior to the procedure. (See "Risk factors and prevention of peritonitis in peritoneal dialysis", section on 'All procedures'.)

General considerations — During the preoperative anesthetic assessment, the following details are determined:

●Type of dialysis (hemodialysis, peritoneal dialysis).

●Frequency of hemodialysis or peritoneal dialysis.

●Date and time of most recent dialysis. Although patients on maintenance hemodialysis will generally be dialyzed 12 to 24 hours prior to elective surgery, those undergoing nonelective surgery may not have received dialysis for up to 72 hours prior to presentation, or longer if there has been a missed treatment. Patients on peritoneal dialysis generally are receiving daily dialysis that may be continued until just before surgery. (See "Medical management of the dialysis patient undergoing surgery", section on 'Routine dialysis prior to surgery'.)

●Type and location of dialysis access.

●Usual fluid intake (may be restricted; therefore, care with perioperative intravenous [IV] fluid volumes is required).

●Usual daily urine output (may be zero).

●"Dry weight" (ie, the target weight). The target weight is commonly established by the outpatient nephrologist as documented on the outpatient dialysis chart. The target weight may not have been achieved after the last dialysis session, or may not be accurate in patients who have recently lost weight due to illness.

●Blood urea nitrogen (BUN) and creatinine concentrations.

●Serum electrolyte concentrations.

●Medications given at dialysis that may not be on usual medication list.

●Name and contact information of patient's nephrologist/dialysis facility.

Hyperkalemia — Hyperkalemia is a potential indication for preoperative dialysis. There are no guidelines that specify a maximum safe level of potassium prior to induction of anesthesia. Decisions regarding treatment of hyperkalemia depend upon the urgency of surgery (ie, whether it is safe to delay surgery for three to four hours to perform dialysis), as well as the likely degree of tissue damage and release of potassium during the planned operation, anticipated blood loss and fluid shifts, chronicity of hyperkalemia, and existing or impending acid-base disturbances that may affect the intraoperative rate of rise of the serum potassium concentration (eg, metabolic acidosis). (See "Medical management of the dialysis patient undergoing surgery", section on 'Hyperkalemia'.)

For patients with chronic hyperkalemia, anesthesia is usually well-tolerated if there are no changes on the electrocardiogram (ECG) and potassium is <6.0 mEq/L. All patients with an elevated serum potassium concentration should have a 12-lead ECG. Hyperkalemia-induced changes in the ECG result from alterations in the transcellular potassium gradient, rather than the absolute serum potassium value. Since dialysis patients often have elevations in total body and intracellular potassium, these transcellular gradients may not be altered with moderate hyperkalemia, accounting for the absence of hyperkalemic changes on the ECG until serum potassium concentrations exceed 6.0 to 6.5 mEq/L [1-5]. However, there is neither an orderly progression of ECG abnormalities in hyperkalemic patients as their potassium rises, nor does the absence of ECG changes preclude the possibility of hyperkalemia-associated cardiac arrest [3-6].

Elective surgery — The serum potassium concentration should be checked on the morning of surgery. If the serum potassium is ≥5.5 mEq/L, we obtain an ECG. For elective surgery, induction of anesthesia in a patient with a serum potassium level <5.5 mEq/L is generally reasonable. If serum potassium is ≥5.5 mEq/L, we generally dialyze the patient. Two to three hours of hemodialysis will typically suffice to prepare a patient for surgery [1]. (See "Medical management of the dialysis patient undergoing surgery", section on 'Elective surgery'.)

Urgent or emergency surgery — For emergency surgery in a patient with potassium ≥5.5 mEq/L, management depends on the following electrocardiographic and surgical factors (see "Medical management of the dialysis patient undergoing surgery", section on 'Nonelective surgery'):

●If there are no ECG changes in an otherwise stable patient, we generally proceed with surgery using continuous intraoperative ECG monitoring and frequent measurement (eg, every 30 minutes) of potassium [7].

●If any ECG features of hyperkalemia are present, we dialyze the patient if feasible, since even one to two hours of hemodialysis is sufficient to reduce total body potassium and serum potassium concentration.

●In a life-threatening surgical situation when dialysis is not feasible (eg, significant hemorrhage), the operation is performed regardless of potassium level and ECG changes. For patients with potassium >6.5 mEq/L or hyperkalemic ECG changes, the anesthesiologist must temporize with medical management of hyperkalemia. Specific treatment for a hyperkalemic emergency includes (algorithm 1 and table 1) (see "Treatment and prevention of hyperkalemia in adults") [6]:

•IV calcium (eg, calcium chloride 500 to 1000 mg) to directly antagonize the cell membrane actions of hyperkalemia. Since hypocalcemia exacerbates potassium-induced cardiotoxicity, ionized calcium levels are monitored, and hypocalcemia is treated. (See "Treatment and prevention of hyperkalemia in adults", section on 'Calcium'.)

•IV insulin (typically given with intravenous glucose) to drive extracellular potassium into cells. (See "Treatment and prevention of hyperkalemia in adults", section on 'Insulin with glucose'.)

•In addition to the measures defined above, bicarbonate therapy 1 to 2 mEq/kg may be administered to raise pH and drive extracellular potassium into cells if severe acute metabolic acidosis is present (ie, pH <7.1 to 7.2). The bicarbonate dose may be repeated if pH remains <7.1 after 30 minutes. Further details are available elsewhere:

-(See "Potassium balance in acid-base disorders", section on 'Metabolic acidosis'.)

-(See "Approach to the adult with metabolic acidosis", section on 'Overview of therapy'.)

-(See "Bicarbonate therapy in lactic acidosis".)

If necessary, continuous renal replacement therapy or hemodialysis can be performed in the operating room if equipment and personnel are available.

Succinylcholine is avoided in patients with hyperkalemia (ie, a serum potassium ≥5.5 mEq/L) due to the potential for increasing the potassium level further and inducing life-threatening arrhythmias preceded by rapidly changing ECG findings [6]. (See 'Neuromuscular blocking agents' below.)

Fluid imbalances — Increased intravascular volume, in particular clinically-significant pulmonary edema, is another potential indication for preoperative dialysis. For urgent or emergency surgery, the risks of mild, moderate, or severe circulatory volume increases are weighed against the risks of delaying surgery.

Optimal volume status prior to surgery is based in part upon estimates of anticipated fluid losses during surgery. Discussions among the surgeon, anesthesiologist, and nephrologist can avoid the following perioperative scenarios:

●If euvolemia or estimated dry weight is not achieved, and/or the patient receives a large volume of fluid during surgery, hypervolemia and possibly pulmonary edema can occur in the immediate postoperative period, necessitating dialysis and possibly noninvasive positive pressure ventilation or controlled mechanical ventilation.

●If too much fluid is removed with preoperative dialysis, risk of intraoperative hypotension due to anesthesia-induced systemic vasodilatation is increased. This may result in significant complications such as thrombosis of the arteriovenous access site.

Typically, the goal of dialysis prior to surgery is to reach an appropriately estimated target weight, although this may differ from the target weight used in the patient's outpatient dialysis setting, as noted above (see 'General considerations' above). Additional factors are considered such as the amount of volume that will be administered or anticipated blood loss during surgery.

Patients with ESKD and chronic heart failure (HF) may benefit from invasive hemodynamic monitoring to more accurately assess and achieve appropriate intravascular volume status. (See "Intraoperative management for noncardiac surgery in patients with heart failure", section on 'Monitoring'.)

Considerations for heparinization — If hemodialysis is performed on the day of surgery:

●For elective or semi-elective cases, perioperative coagulopathy may be avoided by performing heparin-free dialysis on the day of surgery or by waiting for coagulation status to return to normal after hemodialysis with heparin, typically four hours after heparin termination. Peritoneal dialysis does not require heparin administration. (See "Medical management of the dialysis patient undergoing surgery", section on 'Heparin'.)

●For urgent or emergency surgery after recent hemodialysis with use of heparin, protamine can be administered to reverse the effects of heparin. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Reversal'.)

Assessment of comorbidities — Dialysis-dependent patients commonly have multisystem comorbidities that may impact anesthetic and surgical care. These may be mediated by the primary disease process causing renal failure (eg, diabetes and hypertension) and/or by the adverse effects of chronic renal disease and dialysis, or they may be unrelated medical or surgical conditions.

Cardiovascular disease — Cardiovascular disease is particularly common in patients on dialysis. (See "Medical management of the dialysis patient undergoing surgery", section on 'Cardiovascular evaluation'.)

●Venous thrombosis – Thrombosis of central veins, which may make the insertion of central venous access difficult or impossible.

●Coronary artery disease – (See "Medical management of the dialysis patient undergoing surgery", section on 'Cardiovascular evaluation' and "Evaluation of cardiac risk prior to noncardiac surgery" and "Anesthesia for noncardiac surgery in patients with ischemic heart disease".)

●Cerebrovascular disease – (See "Perioperative care of the surgical patient with neurologic disease", section on 'Cerebrovascular disease'.)

●Peripheral vascular disease – Peripheral vascular disease is common in patients with ESKD, particularly those with diabetes. If intra-arterial blood pressure (BP) monitoring is planned, the results of previously performed Doppler vascular studies may be helpful for selection of potential sites for arterial catheter insertion. Intra-arterial access is always avoided in an extremity with a current fistula or graft site.

●Hypertension – (See "Medical management of the dialysis patient undergoing surgery", section on 'Hypertension' and "Anesthesia for patients with hypertension".)

●Heart failure – Chronic HF is common in dialysis patients. It is particularly important to ensure that preoperative dialysis has achieved appropriate volume status prior to surgery, and meticulous attention to intraoperative fluid management is important [8]. (See "Overview of screening and diagnosis of heart disease in patients on dialysis" and "Intraoperative management for noncardiac surgery in patients with heart failure", section on 'Management of fluids and blood products'.)

●Pulmonary hypertension – Pulmonary hypertension is a frequent comorbid condition in patients with ESKD [9,10]. (See "Pulmonary artery catheterization: Indications, contraindications, and complications in adults".)

●Atrial fibrillation – Atrial fibrillation is more prevalent in patients with ESKD than in the general population [11,12]. If possible, preoperative electrolyte disturbances are corrected with dialysis in order to minimize the risk of new-onset or recurrent atrial fibrillation, or other cardiac dysrhythmias. (See "Medical management of the dialysis patient undergoing surgery", section on 'Laboratory testing'.)

Diabetes — We check the glucose level in the preoperative period and treat levels >180 mg/dL then and throughout the perioperative period in all patients with ESKD. Many dialysis-dependent patients have diabetes since this is the most common risk factor for development of ESKD [13]. Even in those without diabetes, glucose intolerance is a feature of uremia. Surgery and general anesthesia typically lead to worsening hyperglycemia due to neuroendocrine stress responses and impairment of insulin secretion (see 'Glucose control' below). Other aspects of perioperative management of blood glucose in dialysis patients are discussed separately. (See "Medical management of the dialysis patient undergoing surgery", section on 'Glycemic control'.)

Pulmonary disease — Pulmonary edema or pleural effusions may be present in ESKD patients with left-ventricular failure, pulmonary hypertension, and/or volume overload. (See "Medical management of the dialysis patient undergoing surgery", section on 'Volume overload'.)

Gastrointestinal disorders — Gastrointestinal (GI) disorders are common among ESKD patients, including gastroparesis associated with uremia and diabetes, with increased risk of pulmonary aspiration. Inflammation of the esophagus, stomach, and duodenum, upper GI bleeding due to uremia-induced platelet dysfunction, and use of heparin during dialysis may be a relative contraindication for TEE monitoring in a patient with severe esophageal or gastric disorders. (See "Unique aspects of gastrointestinal disease in dialysis patients".)

Anemia — Patients with ESKD are often anemic and managed with iron and/or erythropoietin [14]. (See "Treatment of anemia in dialysis patients" and "Medical management of the dialysis patient undergoing surgery", section on 'Anemia status'.)

Although there is variability in individual approaches, we generally do not transfuse stable ESKD patients with hemoglobin >7 g/dL in the absence of a specific indication [15-18]. In particular, transfusion is avoided for patients on a transplant waiting list, if possible, because transfusion-induced sensitization may increase antibody levels and reduce the likelihood of successful eventual renal transplantation [19]. (See 'Management of bleeding diathesis (elective surgery)' below and "Medical management of the dialysis patient undergoing surgery", section on 'Anemia status'.)

Coagulation abnormalities — ESKD patients may have an increased tendency to bleed during surgery due to multiple factors, including platelet inhibition by antiplatelet agents, uremia, anticoagulation for atrial fibrillation, residual heparin used for dialysis, anemia, and increased production of nitric oxide (a vasodilator and platelet function inhibitor) [20]. However, not all uremic patients have a bleeding diathesis and, rarely, some are hypercoagulable [21]. (See "Medical management of the dialysis patient undergoing surgery", section on 'Bleeding diathesis'.)

Management of preoperative antiplatelet agents is similar to that for patients not on dialysis. (See "Perioperative medication management", section on 'Aspirin' and "Perioperative medication management", section on 'Other antiplatelet agents'.)

Management of bleeding diathesis (elective surgery) — For a dialysis patient with preoperative evidence of a bleeding diathesis (eg, significant or prolonged bleeding from the dialysis fistula or the graft insertion site), we obtain standard coagulation tests including prothrombin time (PT), activated partial thromboplastin time (aPTT), international normalized ratio (INR), and platelet count. If available, we also obtain tests of platelet function, although these may not be reliable for diagnosis of uremic platelet dysfunction. (See "Platelet dysfunction in uremia", section on 'Clinical and laboratory manifestations'.)

If platelet dysfunction from uremia is the likely cause of a bleeding diathesis, we perform dialysis and administer desmopressin (dDAVP) in patients undergoing elective or semi-elective procedures when a delay of two to four hours is possible. (See "Medical management of the dialysis patient undergoing surgery", section on 'Bleeding diathesis'.)

●Dialysis – Dialysis may improve platelet function in severely uremic patients. Dialysis is unlikely to improve a bleeding diathesis in a patient who is undergoing routine maintenance dialysis. Ideally, heparin is avoided in patients who are hemodialyzed on the day of surgery, although the effect of heparin is typically gone after four hours [22]. (See 'Considerations for heparinization' above and "Platelet dysfunction in uremia", section on 'Clinical and laboratory manifestations'.)

●Desmopressin – IV dDAVP 0.3 mcg/kg is administered to facilitate platelet aggregation by increasing the release of large von Willebrand factor (vWF) multimers from endothelial cells [23]. Its effect begins within one hour and lasts four to eight hours. Thus, doses of dDAVP should only be administered directly before (but not in the days leading up to) the procedure. Although there is scant evidence regarding the efficacy of dDAVP for prevention or treatment of perioperative bleeding, additional doses may be administered at 12-hour intervals. Tachyphylaxis typically develops after the second or third dose. (See "Platelet dysfunction in uremia", section on 'Clinical and laboratory manifestations'.)

Management of bleeding diathesis (nonelective surgery)

●Uremic patient with excessive bleeding/bleeding risk – If emergency surgery is required in a dialysis patient with evidence of uremic platelet dysfunction, and there is not adequate time for dialysis, we administer IV dDAVP 0.3 mcg/kg. If heparin was recently used, we also administer an appropriate dose of protamine. (See "Medical management of the dialysis patient undergoing surgery", section on 'Heparin' and "Heparin and LMW heparin: Dosing and adverse effects", section on 'Reversal'.)

●Uremic patient with active bleeding – In addition to the therapy above, if a uremic patient is actively bleeding immediately before or during emergency surgery, we administer platelets (1 apheresis unit or 6 units of pooled platelets) even in the absence of thrombocytopenia. However, platelet therapy is only transiently effective for uremic platelet dysfunction since the transfused platelets will be rendered dysfunctional by the uremic milieu, with variable rapidity. Thus, if active bleeding persists, we administer 10 units of cryoprecipitate in addition to platelet transfusion. Cryoprecipitate enhances platelet aggregation by increasing factor VIII:von Willebrand multimers and/or fibrinogen level. (See "Medical management of the dialysis patient undergoing surgery", section on 'Bleeding diathesis' and "Platelet dysfunction in uremia" and "Platelet transfusion: Indications, ordering, and associated risks", section on 'Platelet function defects'.)

In a uremic patient with persistent bleeding, we transfuse washed leukocyte-reduced red blood cells (RBCs) as necessary to maintain hemoglobin (Hgb) >8 g/dL; some clinicians transfuse RBCs to maintain a higher Hgb ≥10 g/dL. (See "Platelet dysfunction in uremia", section on 'Correction of anemia' and "Practical aspects of red blood cell transfusion in adults: Storage, processing, modifications, and infusion", section on 'Washed red cells'.)

In a uremic patient with active bleeding, we ensure that standard coagulation tests have been sent to the laboratory. If available, we obtain specific laboratory tests of platelet function (eg, platelet aggregometry, PFA-100) and point-of-care testing to diagnose causes of coagulopathy and platelet dysfunction (eg, thromboelastography [TEG], rotational thromboelastometry [ROTEM]). (See "Intraoperative transfusion of blood products in adults", section on 'Standard tests' and "Platelet function testing" and "Intraoperative transfusion of blood products in adults", section on 'Point-of-care tests'.)

Premedication

●Aspiration prophylaxis – Gastroparesis and associated increased risk for pulmonary aspiration during induction of anesthesia is common in patients with ESKD [24,25] (see 'Gastrointestinal disorders' above). If premedication with antacids prior to induction of anesthesia is administered due to high risk for aspiration, sodium citrate should be avoided in a dialysis-dependent patient taking aluminum-containing phosphate binders since citrate increases absorption of aluminum, which is toxic. (See "Rapid sequence induction and intubation (RSII) for anesthesia", section on 'Premedication'.)

●Anxiolytics – If IV midazolam is administered in the immediate preoperative period to treat anxiety, the dose should be reduced and titrated (typically in 0.5 to 1 mg increments). The elimination of midazolam and its main metabolite, a1-hydroxymidazolam, is reduced in patients with ESKD [26,27]. Furthermore, protein binding of midazolam is decreased in ESKD, resulting in an increased plasma level of free midazolam [28,29].

●Opioids – Small doses of IV fentanyl (eg, 25 to 50 mcg) may be administered to treat pain in the immediate preoperative period.

Vascular access — Obtaining intravascular access is often difficult in dialysis patients. Those receiving chronic hemodialysis typically have an arteriovenous fistula or a hemodialysis catheter.

Avoid arteriovenous fistula or graft — Venipuncture and vascular access at current fistula or graft sites must be avoided. Exceptions occur in emergency situations or when alternative vascular access is impossible. If use of an arteriovenous fistula or graft is essential, cannulation by an experienced dialysis nurse or technician is strongly recommended.

During surgery, meticulous care of an existing fistula is essential in order to avoid potential thrombosis. This includes avoiding BP measurements or needle sticks in the extremity with the fistula. Subclavian vein catheters should be avoided on either side. Direct pressure on the fistula should be avoided during patient positioning and throughout the perioperative period.

Potential future fistula sites should also be avoided because of the risk of damaging these blood vessels (eg, veins in the antecubital fossa and the cephalic vein at the wrist, especially in the nondominant arm). (See "Medical management of the dialysis patient undergoing surgery", section on 'Intravenous access'.)

Avoid hemodialysis catheter — Generally, a hemodialysis catheter should not be used for purposes other than dialysis (eg, for drug administration or central venous pressure monitoring). Exceptions occur in emergency situations or when alternative vascular access is impossible [30]. (See "Medical management of the dialysis patient undergoing surgery", section on 'Intravenous access'.)

Sites for intravenous access

●Peripheral venous catheters – Ideally, the veins on the back of the hand are used, and the dominant arm is preferred. Peripherally inserted central catheters (PICC) should be avoided in dialysis patients in order to preserve the superficial veins for future arteriovenous fistulae. If a PICC line is necessary, a tunneled PICC line is preferred. (See "Medical management of the dialysis patient undergoing surgery", section on 'Intravenous access'.)

●Central venous catheters – Central venous cannulation may be difficult, particularly if hemodialysis catheters have previously been inserted into central veins. Any previous studies of the patient's vascular anatomy should be reviewed for relevant information (eg, occluded internal jugular, subclavian, or femoral veins).

We advocate and use ultrasound guidance during insertion of central venous catheters, particularly in the internal jugular vein location and in sites where the patient has a history of prior vascular instrumentation or venous thrombosis. Compared with the anatomic landmark approach, ultrasound guidance has been shown to result in a higher overall successful cannulation rate and a decreased rate of arterial puncture or pneumothorax [31-33]. (See "Principles of ultrasound-guided venous access" and "Overview of central venous access in adults" and "Medical management of the dialysis patient undergoing surgery", section on 'Intravenous access'.)

A new central line should not be placed on the same side as an existing hemodialysis catheter. Also, placement of a central line in a subclavian vein is avoided because of the possibility of inducing subclavian stenosis that may preclude subsequent creation of an arteriovenous fistula in the arm on that side.

INTRAOPERATIVE ANESTHETIC MANAGEMENT

Altered pharmacokinetics — Metabolism and elimination of most anesthetic drugs may be delayed in end-stage kidney disease (ESKD), due to impairment of glomerular filtration and renal tubular function, leading to accumulation of the drugs and their metabolites. Also, the volume of distribution and degree of plasma protein binding of anesthetic drugs may be altered, resulting in higher-than-expected plasma concentrations.

Local anesthesia with monitored anesthesia care — Monitored anesthesia care (MAC) with or without local anesthesia is often selected for dialysis-dependent patients, with administration of intravenous (IV) sedative, anxiolytic, or analgesic medications if appropriate. These medications cause respiratory depression and hypotension in a dose-dependent manner (table 2). In general, medications with rapid onset and short duration of action are preferred for sedation/analgesia during MAC to allow rapid titration of effects and quick recovery. In patients with ESKD, doses should be reduced and carefully titrated to effect since these agents may have delayed metabolism, and variability in alterations in volume of distribution and degree of plasma protein binding as well as excretion. (See "Monitored anesthesia care in adults", section on 'Drugs used for sedation and analgesia for monitored anesthesia care'.)

Regional anesthesia — When appropriate, we offer a regional anesthetic technique such as a peripheral nerve block or a neuraxial (spinal or epidural) anesthetic. However, neuraxial anesthetic techniques, paravertebral blocks, and deep plexus blocks (eg, lumbar plexus) are avoided in patients with coagulopathy, as discussed separately. (See 'Coagulation abnormalities' above and "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication".)

Advantages and disadvantages

●Advantages

•Avoidance of a potentially hazardous general anesthetic in a dialysis patient with multiple comorbidities.

•Avoidance of the need to administer multiple IV anesthetic agents that may have delayed metabolism and excretion.

•Provision of superior postoperative analgesia with reduced requirements for systemic analgesic agents (particularly opioids).

•Possible improvements in patency and lower failure rates for vascular access to achieve dialysis [34-36].

●Disadvantages

•Pre-existence of a bleeding diathesis in many dialysis-dependent patients. Hence, a coagulation profile (eg, international normalized ratio [INR] and partial thromboplastin time [PTT], as well as platelet number) is checked for normality before beginning any neuraxial anesthetic procedure. (See 'Coagulation abnormalities' above.)

•Residual anticoagulation after heparin administration may be present in patients up to four hours after hemodialysis. (See 'Considerations for heparinization' above.)

•For patients receiving heparin during postoperative dialysis, the coagulation profile should be checked prior to removal of a neuraxial catheter placed to provide postoperative analgesia [37-39]. (See "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication".)

•Possible low serum bicarbonate levels, which may slow the onset of action of local anesthetic drugs in patients with ESKD [30].

•Possible shortened duration of local anesthetic action due to reduced protein binding [30].

General anesthesia

Induction

Anesthetic induction agents — Generally, we induce general anesthesia with propofol 1 to 2.5 mg/kg because the pharmacokinetic and pharmacodynamic responses to this agent are not markedly altered by ESKD [40,41]. (See "General anesthesia: Intravenous induction agents", section on 'Propofol'.)

The induction dose of propofol should be reduced (eg, 1 to 2 mg/kg) and titrated carefully in dialysis-dependent patients who are older adults, have known coexisting heart failure (HF), or may be hypovolemic [42-44]. In such patients, a standard induction dose of propofol administered as a bolus may result in profound hypotension due to venous and arterial dilation, as well as decreased myocardial contractility. (See "General anesthesia: Intravenous induction agents", section on 'Dosing considerations' and "General anesthesia: Intravenous induction agents", section on 'Disadvantages and adverse effects'.)

Neuromuscular blocking agents — Properties of neuromuscular blocking agents are noted in the table (table 3).

●Succinylcholine – In patients requiring rapid sequence induction and intubation (RSII) due to risk of aspiration, succinylcholine (SCh) can be used safely as the neuromuscular blocking agent (NMBA) to facilitate laryngoscopy if the potassium concentration is <5.5 mEq/L and there are no electrocardiographic (ECG) changes [45]. As in healthy patients, a transient potassium increase of approximately 0.5 to 1 mEq/L is observed after SCh administration, but this hyperkalemic response is not exaggerated in ESKD patients. Patients with ESKD have reduced levels of plasma cholinesterase, the enzyme that metabolizes SCh. Hence, the neuromuscular block caused by SCh may be prolonged [46]. Further details are available in separate topics. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Succinylcholine' and "Rapid sequence induction and intubation (RSII) for anesthesia", section on 'Succinylcholine'.)

●Nondepolarizing neuromuscular blocking agents – For patients with ESKD with potassium ≥5.5 mEq/L, we typically avoid SCh and use either the nondepolarizing NMBA atracurium (0.5 mg/kg) or cisatracurium (0.15 mg/kg) to facilitate laryngoscopy if the patient does not require RSII. Elimination of these agents is independent of renal function [47-49]. However, because of their slow onset (three to four minutes for atracurium and five to seven minutes for cisatracurium), atracurium and cisatracurium are not ideal for patients who require RSII due to aspiration risk. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Atracurium' and "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Cisatracurium'.)

Therefore, we use a relatively large dose of rocuronium (1 mg/kg) rather than SCh in patients who require RSII if potassium is ≥5.5 mEq/L. Although rocuronium is primarily eliminated by direct liver uptake and excretion in bile, some is excreted renally, and neuromuscular blockade may be markedly prolonged after administration of a large intubating dose of rocuronium. However, neuromuscular blockade can be reversed with use of sugammadex after administration of rocuronium, and the sugammadex/rocuronium complex can be removed by dialysis if necessary. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Sugammadex'.)

Remifentanul intubation techniques — A remifentanil intubation technique can be used to facilitate laryngoscopy while avoiding SCh as well as any nondepolarizing NMBA. Propofol 1 to 2 mg/kg is administered followed by a relatively high dose of the ultrashort-acting opioid remifentanil (eg, 2 to 3 mcg/kg), typically providing good intubating conditions in approximately two minutes. We administer ephedrine 10 mg together with these doses of propofol and remifentanil to minimize the profound bradycardia and hypotension that may otherwise result from combined large doses of both agents. (See "Rapid sequence induction and intubation (RSII) for anesthesia", section on 'Remifentanil intubation'.)

Maintenance

Inhalation versus total intravenous anesthesia — We employ either an inhalation-based technique with a potent volatile agent (isoflurane, sevoflurane, or desflurane) with or without nitrous oxide (N₂O) to maintain general anesthesia or total intravenous anesthesia (TIVA). (See "Inhalation anesthetic agents: Clinical effects and uses".)

With inhalation-based techniques, concerns have been expressed regarding the theoretical renal toxicity of sevoflurane due to its inorganic fluoride ion metabolite [50] and formation of a substance known as "Compound A," particularly when low fresh gas flows are used and when higher temperatures are present in the breathing circuit [51-53]. However, sevoflurane has been used safely in patients with chronic stable renal insufficiency and in dialysis patients [54-64]. (See "Inhalation anesthetic agents: Clinical effects and uses", section on 'Sevoflurane'.)

TIVA is an acceptable alternative technique in patients with ESKD. We typically use a continuous IV infusion of both a hypnotic agent (usually propofol at 50 to 150 mcg/kg/minute) and a short-acting opioid (usually remifentanil at 0.015 to 1 mcg/kg/minute) to maintain TIVA. (See "Maintenance of general anesthesia: Overview", section on 'Total intravenous anesthesia'.)

Opioids — We carefully titrate opioids according to individual patient needs in order to avoid postoperative respiratory depression.

●Short-acting opioids – Generally, the pharmacokinetic and pharmacodynamic responses to short-acting opioids (eg, fentanyl, remifentanil, and sufentanil) are not affected by ESKD, although interindividual variability exists [65-68]. Also, acute alkalinization induced by hemodialysis may increase distribution of opioids across the blood-brain barrier into cerebrospinal fluid (CSF). Thus, it is particularly important to monitor for perioperative respiratory depression [68]. (See "Perioperative uses of intravenous opioids in adults: General considerations".)

•Fentanyl – We often use fentanyl in ESKD patients. Fentanyl is predominantly metabolized in the liver to norfentanyl, an inactive metabolite. It has a short redistribution phase and its free fraction is not different in patients with ESKD compared with normal patients. (See "Perioperative uses of intravenous opioids: Specific agents", section on 'Fentanyl'.)

•Sufentanil – Sufentanil is approximately 5 to 10 times more potent than its parent drug fentanyl and may be used as an alternative. Sufentanil is metabolized in the liver and the small intestine. There are no overall differences in half-life and clearance of sufentanil in ESKD patients compared with healthy controls; however, there is more variability in those with ESKD [69]. (See "Perioperative uses of intravenous opioids: Specific agents", section on 'Sufentanil'.)

•Remifentanil – We use remifentanil via infusion without dose adjustment for appropriate surgical procedures in ESKD patients [30]. Remifentanil is rapidly broken down by nonspecific plasma and tissue esterases; thus, accumulation does not occur regardless of duration of administration. Remifentanil has a predictable offset of action, with a context-sensitive half-time that is consistent in normal patients and in those with ESKD [70]. (See "Perioperative uses of intravenous opioids: Specific agents", section on 'Remifentanil'.)

●Long-acting opioids – We typically avoid long-acting opioids in patients with ESKD in the perioperative period. However, when necessary, we prefer to use hydromorphone, buprenorphine, or methadone, and avoid tramadol or morphine. Detailed discussions of these agents in patients with ESKD are presented elsewhere. (See "Medical management of the dialysis patient undergoing surgery", section on 'Opiates' and "Management of chronic pain in advanced chronic kidney disease", section on 'Pharmacologic treatment of advanced CKD (eGFR <30 mL/min/1.73 m2)' and "Management of chronic pain in advanced chronic kidney disease", section on 'Opioids'.)

Neuromuscular blocking agents — We monitor the degree of neuromuscular blockade after administration of an NMBA. There may be interindividual variability in patients with ESKD in their pharmacokinetic and pharmacodynamic responses to muscle relaxants due to coexisting acidosis and alterations in the volume of distribution (table 3). (See "Clinical use of neuromuscular blocking agents in anesthesia".)

●Specific agents – We prefer atracurium (0.5 mg/kg) or cisatracurium (0.15 mg/kg) when a nondepolarizing NMBA is selected during the maintenance phase of anesthesia since their elimination is independent of renal function [47-49].

•Atracurium – Atracurium is an intermediate-acting benzylisoquinolinium NMBA that is metabolized through nonspecific plasma esterase-mediated hydrolysis and a nonenzymatic, pH- and temperature-dependent degradation called Hofmann elimination. Metabolism is essentially independent of hepatic and renal function, and atracurium has no active metabolites. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Atracurium'.)

•Cisatracurium – Cisatracurium is the cis-isomer of atracurium and is four times more potent than atracurium. In contrast with atracurium, it does not cause histamine release. Cisatracurium is also primarily metabolized through Hofmann elimination and has no active metabolites. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Cisatracurium'.)

•Mivacurium – Mivacurium (0.07 to 0.25 mg/kg) may be used in ESKD patients undergoing surgical procedures that are of moderate duration (one or more hours). Mivacurium is hydrolyzed by plasma cholinesterase, like SCh, and does not have active metabolites. Its usual duration of action is 15 to 20 minutes, but recovery may be slower in patients with ESKD due to reduced plasma cholinesterase activity [71]. Reversal (antagonism) of mivacurium with either neostigmine or edrophonium is faster than spontaneous recovery [72]. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Mivacurium'.)

•Rocuronium – Rocuronium (0.6 mg/kg) is sometimes used in patients with ESKD. Rocuronium has a faster onset than other nondepolarizing NMBAs. Thus, rocuronium may be used at higher doses as an alternative to succinylcholine for RSII. It is excreted mostly through the biliary route, but some is excreted renally such that clearance is reduced by 33 to 39 percent in patients with ESKD. Its metabolism results in 17-desacetyl-rocuronium, a compound that has 20 percent the activity of the parent compound. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Rocuronium'.)

•Vecuronium – Vecuronium (0.1 mg/kg) may be used in patients with ESKD only in surgical procedures expected to last several hours. It is not used as an alternative to succinylcholine for RSII because of slower onset time compared with rocuronium. It is excreted both through the biliary and renal routes; thus, neuromuscular blockade is somewhat prolonged and maintenance doses should be carefully guided by neuromuscular monitoring [73,74]. Vecuronium is metabolized to 3-desacetyl-vecuronium, a compound that has 60 percent the activity of the parent compound. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Vecuronium'.)

•Pancuronium – We avoid pancuronium in dialysis-dependent patients [75]. This long-acting muscle relaxant and its active metabolite undergo primarily renal excretion. Since clearance is significantly decreased in patients with ESKD, the paralytic effect may be extremely prolonged. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Pancuronium'.)

●Reversal of nondepolarizing NMBAs – (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Reversal of neuromuscular block'.)

•Neostigmine – We use neostigmine in dialysis patients when reversal of paralysis is required after administration of a nondepolarizing NMBA. With typical single-dose administration at the end of surgery, neostigmine pharmacokinetics are not different from patients with normal renal function [76].

•Sugammadex – Sugammadex is a chelating agent that encapsulates rocuronium or vecuronium in order to rapidly reverse their neuromuscular blocking effects [77]. If high-dose rocuronium is used and reversed with sugammadex in a patient with ESKD, the sugammadex-rocuronium complexes are retained in the body longer than in healthy patients before eventually being excreted by the kidneys. However, the clinical significance of this is not clear [78,79]. Also, the sugammadex-rocuronium complex is removed by dialysis [80].

Fluid management — Management of IV fluid administration may be challenging. Hypervolemia may lead to pulmonary edema, while hypovolemia may cause hemodynamic instability [30]. (See "Intraoperative fluid management", section on 'Consequences of intravascular volume derangements'.)

We administer fluids in 500-mL infusion bags with a micro-dripper to avoid fluid overload, unless large fluid shifts and/or a large volume of blood loss is likely.

●Crystalloids – Centers vary in their selection of IV solutions for dialysis patients. Selections include balanced electrolyte solutions, normal saline (ie, 0.9 percent saline), or normal saline together with a balanced electrolyte solution in a 50-50 combination. (See "Intraoperative fluid management", section on 'Crystalloid solutions'.)

Patients with ESKD are at risk for development of hyperkalemia when nil per os (NPO) and receiving IV fluids without glucose. Although crystalloid solution with or without 5 percent dextrose may be selected, glucose-containing solutions are avoided in patients with hyperglycemia or if hypokalemia is present. Blood glucose levels are monitored when dextrose-containing solutions are administered.

Administration of large volumes of normal saline may result in hyperchloremic metabolic acidosis compared with administration of balanced electrolyte solutions [81,82]. (See "Intraoperative fluid management", section on 'Crystalloid solutions'.)

●Colloids – In the United States, 5% albumin is typically administered if urgent and significant volume expansion becomes necessary and packed red blood cell (RBC) transfusion is not available or is not indicated; other colloids (eg, gelofusine) may be used in this situation in other countries. (See "Intraoperative fluid management", section on 'Colloid solutions'.)

●Blood – Perioperative blood transfusion is avoided, if possible. However, in patients with ongoing surgical bleeding or a hemoglobin <7 g/dL, transfusion of RBCs is often clinically indicated. Potassium should be checked after transfusion since hyperkalemia may develop in an anuric patient (see 'Fluid and electrolyte management' below). Further discussion of transfusion and alternatives to transfusion in patients with chronic kidney disease is presented separately. (See 'Anemia' above and "Treatment of anemia in dialysis patients".)

Glucose control — We maintain blood glucose <180 mg/dL (<10 mmol/L) throughout the perioperative period in both diabetic and nondiabetic patients. Either hyperglycemia or hypoglycemia may occur, particularly in dialysis patients with type 1 diabetes.

If insulin is administered, the serum glucose level is checked one hour later, or every 30 to 60 minutes when a perioperative insulin infusion is used [83]. It is particularly important to avoid hypoglycemic episodes. Although hyperglycemia is associated with an approximately two- to fourfold increased risk of a myocardial ischemic event in noncardiac surgery [84,85], attempts to tightly control glucose (81 to 108 mg/dL [4.5 to 6 mmol/L]) may cause harm due to inadvertent hypoglycemia [86]. (See "Medical management of the dialysis patient undergoing surgery", section on 'Glycemic control' and "Glycemic control in critical illness".)

POSTOPERATIVE ANESTHETIC CARE — Most dialysis-dependent patients can be admitted to a post-anesthesia care unit (PACU) then return home after an outpatient procedure, or to a regular surgical ward after an inpatient procedure. Admission to a high dependency or intensive care unit (ICU) is appropriate for dialysis patients who are hemodynamically unstable after major surgical procedures and for those with recognized perioperative complications.

Fluid and electrolyte management — Fluid management and dialysis in the postoperative period requires consultation with a nephrologist. Blood urea nitrogen (BUN), creatinine, and electrolyte levels should be checked in the early postoperative period.

Ideally, dialysis should be delayed until the risk of surgery-induced fluid shifts and hemorrhage has diminished. If the patient is on hemodialysis, heparinization of the circuit may be reduced or omitted to avoid postoperative bleeding [30]. In some hemodynamically unstable patients, continuous renal replacement therapies (CRRT) may be used instead of hemodialysis in the postoperative period. (See "Prescription of continuous renal replacement therapy in acute kidney injury in adults" and "Continuous kidney replacement therapy in acute kidney injury".)

Pain management in the immediate postoperative period

●General considerations – We use a multimodal approach to postoperative analgesia and avoid long-acting opioids due to potential accumulation of metabolites in patients with end-stage kidney disease (ESKD). (See 'Opioids' above.)

Meperidine is avoided (eg, for treatment of shivering). The renally excreted active metabolite is normeperidine, which will accumulate and may cause respiratory depression as well as adverse neuroexcitatory effects (eg, myoclonic activity and/or seizures) in patients with ESKD [87]. (See "Management of chronic pain in advanced chronic kidney disease", section on 'Drugs that should be avoided' and "Perioperative temperature management", section on 'Shivering'.)

●Regional anesthetic techniques – We use combinations of regional anesthetic techniques or wound infiltration with local anesthetic agents, as well as nonopioid analgesics, to control postoperative pain without an opioid. (See 'Regional anesthesia' above and "Management of acute perioperative pain", section on 'Local anesthesia'.)

●Nonopioid analgesics

•Acetaminophen – Acetaminophen may be used in dialysis patients without dose modification since it is metabolized by the liver [88]. (See "Management of acute perioperative pain", section on 'Intravenous acetaminophen' and "Management of acute perioperative pain", section on 'Acetaminophen'.)

•Nonsteroidal antiinflammatory drugs – We avoid nonsteroidal antiinflammatory drugs (NSAIDs) in patients who have residual kidney function (eg, patients on chronic peritoneal dialysis who still have some daily urine output or patients who have initiated hemodialysis within 6 to 12 months). This is because of concerns regarding worsening of any residual renal function. (See "NSAIDs: Acute kidney injury".)

Also, NSAIDs may further increase bleeding risk in ESKD patients with a bleeding diathesis because of their effects on platelet function, and they may further increase risk of upper gastrointestinal (GI) bleeding due to their effects on gastrointestinal mucosa. Occasionally, an NSAID may be considered for an individual dialysis patient but only for a limited duration [30,88]. (See 'Coagulation abnormalities' above and 'Gastrointestinal disorders' above.)

●Patient-controlled analgesia with opioids – Adequate pain control may not be easily achieved without opioids after many surgical procedures. For patients requiring short-term opioid administration to control postoperative pain, we use a patient-controlled analgesia (PCA) regimen with fentanyl [30]. (See 'Opioids' above and "Management of acute perioperative pain", section on 'Patient-controlled analgesia'.)

SUMMARY AND RECOMMENDATIONS

●Preoperative considerations for dialysis patients include type of dialysis (hemodialysis, peritoneal dialysis), date and time of most recent dialysis, type and location of dialysis access, usual fluid intake, urine output (if any), and target weight, as well as laboratory values (blood urea nitrogen [BUN], creatinine, and electrolyte concentration). (See 'General considerations' above.)

●The major indications for preoperative hemodialysis include hyperkalemia and volume overload. We generally dialyze patients with potassium ≥5.5 mEq/L prior to surgery. Our approach to hyperkalemic patients who require emergency surgery is defined above. (See 'Hyperkalemia' above and 'Fluid imbalances' above.)

●Preoperative bleeding diathesis may be present in a uremic patient or if heparin was recently administered for dialysis. (See 'Coagulation abnormalities' above.)

•If bleeding is due to uremia and surgery can be delayed for two to four hours, we perform dialysis before surgery. We also typically administer intravenous (IV) desmopressin (dDAVP; 0.3 mcg/kg) to such patients. If dialysis cannot be performed prior to surgery, we administer IV dDAVP. If there is active bleeding we administer platelets (1 apheresis unit or 6 units of pooled platelets) and, if bleeding persists, 10 units of cryoprecipitate. (See 'Management of bleeding diathesis (nonelective surgery)' above.)

•We avoid using heparin with hemodialysis on the day of surgery. If heparin cannot be avoided (or has already been given), surgery is delayed for four hours until heparin effect is terminated. If surgery cannot be delayed, we use protamine to reverse the effects of heparin. (See 'Considerations for heparinization' above.)

●Hemodialysis catheters should not be used for anything but hemodialysis except in an emergency or where alternative vascular access is impossible. Current and potential arteriovenous fistula sites, as well as peripherally inserted central catheters (PICC), are avoided in order to preserve superficial veins for future arteriovenous fistulae. If the patient has a fistula or graft, the arm with the fistula or graft should not be used for blood pressure (BP) measurements, catheter insertion, or needle sticks. Direct pressure on the fistula should be avoided during patient positioning and throughout the perioperative period. (See 'Vascular access' above.)

●Monitored anesthesia care (MAC) with or without local anesthesia is often selected for a patient on dialysis, with administration of reduced and carefully titrated IV sedative, anxiolytic, or analgesic medications as needed (table 2). (See 'Local anesthesia with monitored anesthesia care' above.)

●When appropriate, we offer a regional anesthetic technique such as a peripheral nerve block or a neuraxial (spinal or epidural) anesthetic. However, neuraxial anesthetic techniques, paravertebral blocks, and deep plexus blocks (eg, lumbar plexus) are avoided in patients with coagulopathy. (See 'Regional anesthesia' above.)

●Metabolism and elimination of most anesthetic drugs may be delayed in patients with end-stage kidney disease (ESKD) due to impairment of glomerular filtration and renal tubular function, leading to accumulation of these drugs and their metabolites. Also, the volume of distribution and degree of plasma protein binding of anesthetic drugs may be altered, resulting in higher-than-expected plasma concentrations. (See 'Altered pharmacokinetics' above.)

●Generally, we induce anesthesia with propofol because patient pharmacokinetic and pharmacodynamic responses are minimally altered by ESKD. Considerations for use of a neuromuscular blocking agent (NMBA) to facilitate laryngoscopy include (see 'Induction' above):

•Succinylcholine (SCh) can be used safely during rapid sequence induction and intubation (RSII) if potassium is <5.5 mEq/L.

•For patients with potassium ≥5.5 mEq/L who do not require RSII, we avoid SCh and use either the nondepolarizing NMBA atracurium (0.5 mg/kg) or cisatracurium (0.15 mg/kg) since their elimination is independent of renal function. We also prefer these agents if an NMBA is necessary during the maintenance phase of anesthesia. (See 'Neuromuscular blocking agents' above.)

•If potassium is ≥5.5 mEq/L in a patient who requires RSII, we use a relatively large dose of rocuronium (1 mg/kg) rather than SCh.

•A remifentanil intubation technique can be used to facilitate laryngoscopy while avoiding SCh as well as any nondepolarizing NMBA.

●We use an inhalation agent or total intravenous anesthesia (TIVA) for maintenance of general anesthesia. For surgical procedures that require a total intravenous anesthesia (TIVA) technique, we use a continuous infusion of both a hypnotic agent (propofol) and a short-acting opioid (remifentanil). (See 'Inhalation versus total intravenous anesthesia' above.)

●We carefully titrate IV short-acting opioids (eg, fentanyl, remifentanil, sufentanil) according to individual patient needs in order to avoid postoperative respiratory depression. We minimize or avoid use of longer-acting opioids. (See 'Opioids' above.)

●We use a multimodal approach to postoperative analgesia, including combinations of nonopioid analgesics, regional anesthetic techniques, and wound infiltration with local anesthetic agents in order to control postoperative pain. If an opioid is required, we use an IV patient-controlled analgesic (PCA) regimen with fentanyl and monitor closely for respiratory depression. (See 'Pain management in the immediate postoperative period' above.)