Chronic kidney disease in the critically ill

Contents

Overview

Chronic kidney disease (CKD) is common in ICU patients, affecting their response to critical illness, their pharmacological management, and their outcomes. CKD predisposes to AKI (acute-on-chronic kidney disease), carries an independently high cardiovascular risk, and alters the pharmacokinetics of many drugs used in critical care. Recognition and adaptation of standard management approaches is essential.

Staging and Epidemiology

CKD is defined as kidney damage or a reduced GFR persisting for more than 3 months. The KDIGO classification uses eGFR and albuminuria to stage severity:

Stage eGFR (mL/min/1.73 m²)
G1 ≥90 (with markers of kidney damage)
G2 60–89
G3a 45–59
G3b 30–44
G4 15–29
G5 <15 (or on dialysis)

Albuminuria categories (A1: <30 mg/g, A2: 30–300 mg/g, A3: >300 mg/g) independently predict cardiovascular and renal outcomes at each GFR stage.

Approximately 10–15% of ICU patients have pre-existing CKD. CKD significantly increases the risk of AKI during critical illness and worsens short- and long-term mortality.

Pathophysiology Relevant to the ICU

Uraemic toxin accumulation: Reduced excretion of urea, creatinine, phosphate, and other solutes. Clinically significant manifestations include uraemic pericarditis, uraemic encephalopathy (asterixis, altered consciousness, seizures), and platelet dysfunction causing bleeding tendency.

Reduced acid excretion and bicarbonate regeneration: CKD causes a normal anion gap (hyperchloraemic) metabolic acidosis at earlier stages, progressing to a high anion gap acidosis in advanced CKD as organic acid excretion fails. In critical illness, the metabolic acidosis from CKD compounds lactic acidosis and other superimposed disturbances.

Fluid and sodium homeostasis: As GFR falls, the ability to excrete sodium and water becomes impaired. CKD patients are at high risk of volume overload with standard fluid resuscitation volumes, and are intolerant of the positive fluid balances that are common in critical illness.

Anaemia: CKD causes anaemia from reduced erythropoietin (EPO) production, chronic inflammation, iron deficiency, and shortened red cell survival. Haemoglobin targets in CKD are typically 100–120 g/L; higher targets increase cardiovascular risk (demonstrated in CREATE and TREAT trials). Transfusion thresholds in critical illness should not be altered significantly by CKD status alone.

Cardiovascular disease: CKD accelerates atherosclerosis, causes left ventricular hypertrophy, and predisposes to arrhythmias. The majority of CKD patients die from cardiovascular causes rather than from progression to end-stage renal failure. ICU patients with CKD have a high risk of myocardial ischaemia.

Drug Prescribing in CKD

Antibiotics: Most antibiotics require dose or interval adjustment in CKD. Beta-lactams, aminoglycosides, vancomycin, and carbapenems are primarily renally excreted. Failure to adjust doses causes accumulation and toxicity. Aminoglycosides should be used with caution; if essential, single-dose therapy with drug level monitoring is preferable to prolonged courses.

Opioids: Morphine-6-glucuronide (M6G), the active metabolite of morphine, accumulates in renal impairment and causes prolonged sedation and respiratory depression. Oxycodone also has active metabolites that accumulate. Fentanyl and alfentanil are preferred in CKD as they do not produce clinically significant active metabolites.

Low molecular weight heparin (LMWH): Cleared renally; accumulates in CKD (eGFR <30 mL/min/1.73 m²). Requires dose reduction or use of unfractionated heparin with anti-Xa monitoring. Enoxaparin dose should be halved for VTE prophylaxis when eGFR <30.

NSAIDs: Reduce renal blood flow by inhibiting prostaglandin-mediated afferent arteriolar vasodilation. In CKD, prostaglandins are critical for maintaining GFR; NSAIDs cause acute deterioration in renal function and should be avoided.

Contrast media: Iodinated IV contrast can cause contrast-induced nephropathy (CIN), particularly in CKD G3–5, diabetes, and concurrent nephrotoxin use. Risk mitigation: use the lowest possible volume of iso-osmolar contrast, ensure adequate hydration, hold nephrotoxic drugs. Routine N-acetylcysteine prophylaxis is no longer recommended based on current evidence. The risk of withholding contrast for a diagnostic CT when clinically indicated often outweighs the risk of CIN.

Drug dialysability: Many drugs are removed by haemofiltration or haemodialysis. Clinically important examples: vancomycin (supplemental doses required post-HD), meropenem, piperacillin-tazobactam (dose after HD session). Check drug data sheets or pharmacy.

Electrolyte and Metabolic Disturbances

Hyperkalaemia: The most dangerous electrolyte complication of CKD in the ICU. Reduced renal potassium excretion compounded by metabolic acidosis (transcellular shift), haemolysis, rhabdomyolysis, and potassium-containing fluids. Management: ECG monitoring, calcium gluconate for membrane stabilisation, insulin-dextrose, salbutamol, resonium (slow onset), loop diuretics, and RRT if refractory.

Hyperphosphataemia: Accumulation of phosphate causes secondary hyperparathyroidism, metastatic calcification, and renal osteodystrophy. In the ICU, severe hyperphosphataemia can also worsen metabolic acidosis. Managed with phosphate binders (calcium carbonate or sevelamer with meals) and dietary restriction.

Hypocalcaemia: Reduced renal activation of vitamin D impairs intestinal calcium absorption, and hyperphosphataemia causes calcium precipitation. Symptomatic hypocalcaemia (tetany, seizures, prolonged QT) requires IV calcium gluconate. Correction of hyperphosphataemia is equally important to prevent further precipitation.

Metabolic acidosis: Manage the underlying critical illness and support with sodium bicarbonate if pH below 7.2 or bicarbonate below 15 mmol/L, particularly if delaying or exacerbating other treatments. RRT provides effective acid clearance when medical management is insufficient.

Haemodynamic Management

Patients with CKD frequently have established hypertension and a rightward shift of the renal autoregulation curve. Standard MAP targets of 65 mmHg may be insufficient to maintain adequate renal perfusion pressure in a patient whose kidneys are accustomed to higher mean pressures.

Target MAP of 75–80 mmHg is generally recommended for CKD patients in septic shock, based on subgroup analyses from the SEPSISPAM trial (Asfar et al., NEJM 2014) showing reduced need for RRT in patients with chronic hypertension randomised to a higher MAP target.

Fluid management must be conservative. Avoid large positive fluid balances; reassess fluid requirements at each clinical decision point. Dynamic markers of fluid responsiveness (stroke volume variation, pulse pressure variation) are preferable to static filling pressures. Early consideration of renal replacement therapy for fluid removal (ultrafiltration) in patients who become fluid-overloaded without a clinical need for ongoing resuscitation.

Renal Replacement Therapy

CKD patients admitted in AKI-on-CKD may require RRT earlier than equivalent patients without background CKD, as they lack the physiological reserve to tolerate the same degree of metabolic derangement. Standard RRT indications apply (refractory hyperkalaemia, pulmonary oedema, severe acidosis, uraemic complications).

Patients who were already established on intermittent haemodialysis pre-admission should continue on the most appropriate RRT modality for their haemodynamic state — continuous RRT is preferred in haemodynamically unstable patients.

Recovery of renal function to baseline is possible in acute-on-CKD but may be slower and less complete than in patients with no prior CKD. Discussions about long-term renal outcomes should involve nephrology early.

Viva Questions

How does CKD alter the pharmacokinetics of drugs commonly used in the ICU?

CKD reduces renal drug clearance, causing accumulation of renally cleared drugs and their active metabolites. For antibiotics, this typically requires dose reduction or interval extension: aminoglycosides, beta-lactams, vancomycin, and carbapenems all require adjustment below an eGFR of approximately 30–50 mL/min depending on the agent. Morphine accumulates its active metabolite M6G, causing prolonged sedation and respiratory depression — fentanyl is preferred. LMWH accumulates in CKD and should be dose-adjusted or replaced with UFH in significant renal impairment. The volume of distribution may also be altered in CKD due to hypoalbuminaemia (reducing protein binding of acidic drugs, increasing free fraction) and fluid overload. In addition, many drugs that are cleared by dialysis require supplemental dosing after each RRT session to maintain therapeutic levels.

What electrolyte and metabolic disturbances occur in advanced CKD and how do you manage them in the critically ill patient?

The principal metabolic disturbances are hyperkalaemia, hyperphosphataemia, hypocalcaemia, and metabolic acidosis. Hyperkalaemia is the most immediately dangerous: management begins with ECG assessment and, if changes are present, immediate calcium gluconate IV for membrane stabilisation, followed by insulin-dextrose and salbutamol to drive potassium intracellularly. Resonium and loop diuretics provide slower potassium removal; RRT is indicated when the above measures fail. Hyperphosphataemia is managed with phosphate binders given with meals and dietary restriction. Hypocalcaemia secondary to reduced vitamin D activation requires supplementation with alfacalcidol or calcitriol rather than plain vitamin D (the renal activation step is impaired). Symptomatic hypocalcaemia is treated with IV calcium gluconate. Metabolic acidosis may require sodium bicarbonate infusion to maintain pH above 7.2, and RRT is effective at clearing acid when the clinical situation demands it.

What MAP target would you aim for in a patient with known CKD and why?

In a patient with established CKD and background hypertension, I would target a MAP of 75–80 mmHg rather than the standard 65 mmHg threshold used in normotensive patients. The rationale comes from the subgroup analysis of the SEPSISPAM trial, which randomised patients with septic shock to a MAP target of 80–85 vs 65–70 mmHg. In the overall population there was no mortality difference, but in the subgroup of patients with chronic hypertension, the higher MAP target was associated with a lower rate of AKI requiring renal replacement therapy. The biological explanation is autoregulation: in patients with long-standing hypertension, the renal autoregulation curve is shifted to the right, so adequate perfusion pressure is only achieved at higher MAPs. Standard targets may lie below the lower limit of autoregulation in these patients, leading to pressure-dependent reductions in renal blood flow and GFR. In practice, I would use vasopressors to achieve the higher target and monitor urine output, creatinine trend, and fluid balance closely as surrogate markers of renal perfusion adequacy.