The risk of developing chronic kidney disease (CKD) increases with age and glomerular filtration rates (GFR) <60 mL/min per 1.73 m2 are relatively common. Almost a quarter of hospital inpatients suffer acute kidney injury (AKI).
Due to the central role of the kidney in homeostasis, including regulation of circulating blood volume, blood pressure control and effect on plasma electrolyte composition, compromised kidney function can have significant impact on anaesthesia.
Patients with renal disease have an increased incidence of hypertension and other cardiovascular disease. A systematic approach should be followed to identify risks such as coronary artery disease with a view to revascularisation or other optimisation interventions.
Approximately one third of AKI cases in the elderly result from perioperative hypotension with or without hypovolaemia. Preoperative evaluation should include a careful assessment of intravascular fluid status, followed by appropriate measures to achieve euvolaemia.
Anaesthetist should be familiar with the management of electrolyte imbalances, metabolic acidosis and the pharmacokinetics of drugs eliminated by kidneys.
Patients with renal disease are frequently anaemic and are at increased risk of bleeding perioperatively.
The kidney’s position at the centre of homeostasis impacts a number of areas of anaesthetic practice. It has critical roles in the regulation of circulating blood volume, blood pressure and plasma electrolyte composition in addition to functioning as an endocrine organ in the regulation of anaemia and calcium homeostasis. The kidney plays a key role in elimination of a number of metabolic waste products, including urea and excess hydrogen ions. Likewise, a large number of drugs are either eliminated by direct excretion via the kidney or water-soluble metabolites are dependent on renal elimination.
How Common Is Renal Disease?
The risk of developing chronic kidney disease (CKD) increases with age. Having some degree of CKD with glomerular filtration rate <60 mL/min per 1.73 m2 is relatively common with an age standardised prevalence in the UK of 8.5 per cent. A comprehensive review of the evidence on screening for CKD and its management can be found in NICE Clinical Guideline 182 (www.nice.org.uk/guidance/cg182). End-stage renal disease (ESRD) is less common with incidence in the UK in 2013 of 109 per million population. The prevalence of patients on renal replacement therapy was 888 per million of the population, of which 52 per cent had a functioning renal transplant with 42 per cent on haemodialysis and 6 per cent on peritoneal dialysis (www.renalreg.org). In the UK in 2012–13, 209 per million population had acute kidney injury (AKI) sufficiently severe to require renal replacement therapy (Kohle et al., 2015) and many more patients had less severe AKI, typically 15–20 per cent of hospital inpatients.
Normal glomerular filtration rate (GFR) is 90–120 mL/min per 1.73 m2 and from the age of 30 years decreases by approximately 1 mL/min/1.73 m2 per year. In situations where it is essential to have an accurate measure of GFR, clearance of an intravenously administered tracer molecule from blood, usually iohexol or Cr51 labelled EDTA, is measured. This is rarely necessary, but is important in assessment of potential live kidney donors or in the use of cancer chemotherapeutic agents with a narrow therapeutic index and renal elimination. Approaches using creatinine as a biomarker to estimate GFR are much more practical. Cystatin-C has been introduced more recently and may have some benefits over creatinine, although it has not yet been widely adopted. Measured 24-hour creatinine clearance has fallen from use, primarily because patients find well-timed urine collections difficult, resulting in poor accuracy. In a hospital setting with a bladder catheter in situ, timed creatinine clearance actually does give a reasonable measure of GFR. Creatinine is a metabolic by-product from skeletal muscle and steady state is based on a combination of rate of generation which is proportional to muscle mass and the rate of renal excretion which is primarily through glomerular filtration with a small contribution from tubular secretion. An individual with a large muscle mass may have the same serum creatinine concentration as a small emaciated individual with advanced renal impairment. As a consequence, it is impossible to define a normal range for serum creatinine. Some drugs, such as trimethoprim, block tubular secretion with a consequent rise in serum creatinine concentration but no change in GFR. Changes within an individual usually do indicate changes in renal function. The relationship between GFR and serum creatinine is logarithmic, resulting in relatively minor increase in serum creatinine concentration until GFR falls to less than 50 per cent of normal but then changes progressively more with further falls in GFR. The equations derived for estimating GFR (eGFR) factor in an estimate of muscle mass. The Cockcroft-Gault equation uses sex, age and weight, while the more widely used Modification of Diet in Renal Disease (MDRD) and Chronic Kidney Disease – Epidemiology Collaboration (CKD-EPI) equations use standard laboratory parameters. Conventionally GFR is expressed per 1.73 m2 body surface area. GFR calculators based on these equations are widely available for hand-held devices. There are several important caveats in interpreting estimates of GFR. The MDRD equation widely used in laboratory reporting was derived from a population of patients with CKD, most of whom had GFR in the range <60 mL/min. As a consequence, the equation is unreliable at GFRs greater than this, and a report of ≥60 mL/min/1.73 m2 indicates that it is beyond the range of the estimate, not that it is normal. The more recently introduced CKD-EPI equation has better performance across the full GFR range (Levey et al., 2009). These equations have not been validated for use in acute kidney injury.
Both CKD and AKI have been categorised by degree of reduction in GFR. The Kidney Disease: Improving Global Outcomes (KDIGO) classification is the most widely used (Table 8.1). Stage 5 CKD does not always equate to ESRD. In the UK the mean eGFR for patients starting dialysis in 2013 was 8.5 mL/min/1.73 m2 (www.renalreg.org). Whatever the aetiology of CKD, the process eventually becomes progressive with falling GFR. Control of blood pressure, in particular with ACE-inhibitors or angiotensin-II receptor blockers, slows the progression.
|GFR category||GFR (mL/min/1.73 m2)|
Legend: KDIGO = Kidney Disease: Improving Global Outcome, CFR = glomerular filtration rate, mL/min/1.73 m2 millilitre per minute per 1.73 square metre
AKI has also been categorised based on a combination of serum creatinine concentration and urine output (Table 8.2).
|Stage||Serum creatinine||Urine output|
|1||1.5–1.9 times baseline||<0.5 mL/kg/h for 6–12 hours|
|≥27 µmol/L (0.3 mg/dL) increase|
|2||2.0–2.9 times baseline||<0.5 mL/kg/h for ≥12 hours|
|3||3.0 times baseline||<0.3 mL/kg/h for ≥24 hours|
|Increase to ≥354 µmol/L (4.0 mg/dL)||Anuria for ≥12 hours|
|Initiation of renal replacement therapy|
|In patients aged <18 years a decrease in eGFR to <35 mL/min/1.73 m2|
Legend: mL/kg/h = millilitre per kilogram per hour, µmol/L = micromol per litre, mg/dL=milligram per decilitre, eGFR= estimated glomerular filtration rate, mL/min/1.73 m2 millilitre per minute per 1.73 square metre
Small amounts of protein are normally filtered through the glomerulus and are catabolised by the renal tubule. Tubular dysfunction may result in a small amount of proteinuria, but, in general, significant proteinuria indicates glomerular disease. Stick-testing is semi-quantitative and tests for albumin only. Measurement of 24-hour urinary protein has been abandoned for practical reasons and been replaced by measurement of the ratio of albumin or protein in urine to creatinine to correct for changes in urine concentration. The KDIGO classification for CKD now includes the degree of proteinuria (Kidney Disease: (KDIGO Improving Global Outcomes) CKD work group, 2013) (Table 8.3).
|ACR category||Albumin/creatinine ratio (mg/mmol)|
ACR=albumin creatinine ratio, mg = milligram, mmol=millimol
Susceptibility to AKI due to hypovolaemia increases with age and is more common in individuals with pre-existing CKD. Approximately one third of cases of AKI in the elderly result from perioperative hypotension with or without hypovolaemia. Patients with ESRD on dialysis may have some residual urine output that is worth protecting as it may contribute significantly to prevention of chronic fluid overload. Individuals with diabetes mellitus or myeloma are particularly sensitive to volume depletion. Preoperative assessment should include a careful assessment of intravascular fluid status, followed by appropriate measures to achieve euvolaemia. A patient with ESRD and signs of pulmonary oedema may need dialysis preoperatively to remove fluid. Perioperative management in anuric patients can be challenging. These patients are at high risk for cardiovascular events (later in this chapter is more suitable), making hypovolaemia undesirable, and it is self-evidently important to avoid over-filling: it is not easy to get the excess fluid back out again perioperatively. While central venous pressure monitoring can help inform fluid management, central venous cannulation is often difficult due to previous dialysis access. Other monitoring tools such as these based on pulse contour analysis or oesophageal Doppler represent an important advance. Preoperative chest radiographs in patients on haemodialysis are best taken after a dialysis session when they provide an indication of pulmonary vasculature closer to optimal blood volume than in patients prior to dialysis who tend to be fluid overloaded.
The kidney is central to the regulation of systemic blood pressure. Hypertension is common in patients with renal disease due either to chronic salt and water retention or renin production from ischaemic areas of the kidney. Patients on treatment by haemodialysis are usually hypertensive due to fluid overload prior to a dialysis session and normotensive at the end of dialysis following fluid removal. Hypertension is a key risk factor for cardiovascular disease, as discussed later in this chapter. Arterial cannulation to monitor blood pressure may be rendered difficult by vascular calcification or previous surgery to form arteriovenous fistulae for haemodialysis access.
An inverse correlation exists between glomerular filtration rate and cardiovascular risk. Sudden cardiac death in the absence of a myocardial infarct is the most common cause of death in patients with renal failure. Cardiovascular disease in patients with renal disease is complex with atherosclerosis and nonatherosclerotic changes, including endothelial dysfunction, arterial wall thickening and vascular calcification leading to arterial stiffness. Coronary calcification detected by electron beam CT scanning occurs at a younger age in patients with CKD and they have higher calcification scores than the general population. Important non-classical cardiovascular risk factors include abnormalities in calcium, phosphate and parathyroid hormone homeostasis, inflammation, oxidative stress, vitamin D deficiency and anaemia. Recently, urate retention has also been implicated as a cardiovascular risk factor. There is also a high incidence of classical atherosclerotic disease and traditional cardiovascular risk factors; in particular hypertension and diabetes mellitus are common. Heart failure and arrhythmia are more frequent causes of death than occlusive arterial disease.
In a classical study of echocardiographic features of patients starting dialysis, 44 per cent had concentric left ventricular hypertrophy (LVH) and 30 per cent had increased cavity volume (Foley et al., 1995). LVH is likely to be due to pressure overload as a consequence of arterial stiffness and hypertension, whereas dilatation is a consequence of chronic fluid overload. Anaemia is a further key risk factor for LVH which is probably ameliorated by treatment with erythropoietin. Timing of echocardiography in relation to dialysis is important, and cardiac function can be underestimated in volume-loaded patients due to undergo haemodialysis with improved function at optimal blood volume. Cardiac magnetic resonance imaging (MRI) findings are volume independent and demonstrated LVH with preserved systolic function in around 15 per cent of patients. Left ventricular dilatation with impaired systolic function was observed in a further 15 per cent associated with a higher burden of coronary artery disease (Mark et al., 2006). The identification of nephrogenic systemic sclerosis as a complication of gadolinium use as a contrast agent has prevented ongoing use of cardiac MRI in patients with significant renal impairment. Capillary density in hypertrophied myocardium is reduced, resulting in ischaemia even in the absence of occlusive arterial disease. Impaired intra-cardiac conduction predisposes to arrhythmias, which are two- to three-fold more common in dialysis patients than in the general population. Diastolic dysfunction is due to a combination of stiffening of the conduit arteries, resulting in lower diastolic pressure which may in itself impair coronary perfusion and impaired ventricular relaxation. Individuals with diastolic dysfunction are particularly sensitive to changes in blood volume with hypotension associated with excessive fluid loss, e.g. when haemodialysed or pulmonary oedema when fluid overloaded. It is uncertain whether so-called uraemic cardiomyopathy due to the metabolic changes of uraemia is superimposed on these structural factors. Systolic dysfunction is present in around 20 per cent of dialysis patients and diastolic dysfunction in around 50 per cent. Calcification of the mitral and aortic valves is four times more common in dialysis patients than in normal individuals potentially leading to stenosis or regurgitation. Renal transplantation results in improvement, although not normalisation, of most of these parameters.
Diabetes is now one of the most common causes of renal disease, and autonomic neuropathy is a frequent accompaniment to diabetic nephropathy. Uraemia itself predisposes to autonomic neuropathy. Uraemic pericarditis with pericardial effusion is a rare occurrence in patients who have been uraemic for a prolonged period prior to presentation and is unusual in patients on renal replacement therapy.
Assessment of Cardiovascular Risk in Patients with Renal Disease
Approximately 50 per cent of the deaths during the first month after renal transplantation are due to cardiovascular disease. Hence, a major focus of pre-transplant assessment is detection of occult cardiac disease and assessment of cardiovascular risk. The standard approach used in assessing suitability of patients for renal transplantation will be used as a paradigm for cardiovascular assessment in patients with renal disease. While these assessments will have been undertaken routinely for patients on the transplant waiting list, they are not part of routine management of patients with renal disease. The evidence base supporting this approach to screening is weak and there is wide divergence of practice in this area.
Echocardiography quantifies left ventricular function with the caveats noted previously around the volume status of the patient who may have significantly better function with optimisation of blood volume. LVH and valvular heart disease, which is a particular problem due to the tendency to calcify heart valves, are identified.
A risk-based approach to the identification of coronary artery disease with a view to revascularisation or definition of risk is undertaken (Atkinson et al., 2011). In patients who are aged less than 50 years and are not diabetic with no history of angina, previous myocardial or cerebrovascular events or peripheral vascular disease and a normal ECG or ECG changes compatible with isolated LVH, no further investigation is warranted. For higher-risk patients, there are advocates for proceeding to coronary angiography in all with the alternative approach based on functional tests of myocardial ischaemia to identify those at high risk prior to proceeding to coronary angiography. Exercise ECG performs poorly, although it does provide a holistic functional test. This function can be achieved more straightforwardly by a 6-minute walk test where the ability to perform greater than 6 minutes of a Bruce protocol predicts lower perioperative mortality. The choice between stress echocardiography and radionucleotide imaging depends on local expertise and availability. Those with evidence of inducible ischaemia proceed to coronary angiography with revascularisation based on conventional prognostic indications (see chapter on cardiovascular disease). A new 12-lead ECG should be obtained prior to anaesthesia in all patients.
Measures to Reduce Cardiovascular Risk
Beta-blockers have been shown to confer prognostic benefit in haemodialysis patients, and many patients with ESRD will be on long term beta-blockade. While there is some controversy over the detriment of stopping beta-blockers perioperatively, it is probably safest to continue them. Please refer to Chapter 1.
Plasma potassium concentration is tightly regulated in the range 3.5–5.3 mmol/L, largely by adjusting renal excretion. The acceptable range for anaesthesia is generally considered to be 3.5–5.5 mmol/L. Dialysis is the only option for clearance of excess potassium in anuric patients. A key caveat to remember is that immediately after a haemodialysis session, serum potassium concentrations are often below the normal range but rebound quickly within 1–2 hours after a dialysis session. Resist the temptation to treat post-haemodialysis hypokalaemia.
Hyperkalaemia is common due to reduced renal excretion or a shift of potassium from the intracellular to the extracellular compartment due to acidaemia. Depolarising muscle relaxants, in particular suxamethonium, increase plasma potassium through release of intracellular potassium, typically by 0.5 mmol/L. Other causes of perioperative hyperkalaemia include reabsorption of blood and tissue trauma. Potassium containing intravenous fluids, e.g. Hartmann’s solution, should be avoided in patients with ESRD. Problems are unusual with K+<6.0 mmol/L, but most anaesthetists would be reluctant to anaesthetise a patient with serum K+ >5.5 mmol/L. The most important consequence of hyperkalaemia is cardiac arrhythmias, specifically, asystole and ventricular fibrillation. Typical ECG changes are loss or flattening of P-waves, widening of the QRS complex and enlarged ‘tented’ T-waves.