Detailed Notes: Multidisciplinary Guidelines on Renal Replacement Therapy in Intensive Care Medicine
Source: Meersch-Dini et al.,
Critical Care, DOI: 10.1186/s13054-025-05817-6. The guideline is a multidisciplinary German-Austrian consensus document for renal replacement therapy (RRT) in critically ill adults with dialysis-dependent acute kidney injury (AKI). The final open-access article is available through
the PMC version.
1. Purpose and scope
Why the guideline was needed
RRT is commonly used in ICU patients with severe AKI. Important decisions remain variable in practice:
- When to begin RRT
- Which clearance method to use: diffusion, convection, or both
- Whether continuous or intermittent therapy is preferable
- Choice and monitoring of anticoagulation
- Appropriate treatment dose
- Drug dosing during RRT
- When and how to discontinue RRT
How the guideline was developed
- A multidisciplinary panel of 22 experts from 12 German-speaking professional societies was assembled.
- Included intensivists, nephrologists, anesthesiologists, internists, surgeons, and patient representatives.
- Evidence was identified through systematic searches of PubMed, Scopus, and Cochrane databases.
- Recommendations were developed using:
- PICO clinical questions
- Modified Delphi process
- Nominal group and consensus conferences
- GRADE framework
- AGREE II guideline methodology
- The panel produced 73 statements/recommendations, with 47 key recommendations summarized in the main paper.
Core message
RRT decisions should be based on the whole clinical picture, not on one laboratory value, one AKI stage, one biomarker, or one rigid timing threshold.
2. Starting RRT in ICU AKI
2.1 Absolute indications: start immediately
RRT should be initiated immediately if there are life-threatening disturbances in:
- Fluid balance
- Severe fluid overload, especially pulmonary edema with hypoxemia or respiratory compromise
- Electrolyte balance
- Especially severe or refractory hyperkalemia
- Acid-base balance
- Severe metabolic acidosis not adequately controlled with medical treatment
- Uremia
- Clinically significant uremic complications, for example encephalopathy, pericarditis, or severe symptoms attributable to kidney failure
Key principle
The document uses the classic emergency framework of hypervolemia, hyperkalemia, acidosis, and uremia. In such circumstances, delaying RRT to obtain further biomarkers or conduct a furosemide stress test is inappropriate.
2.2 Relative indications: individualize the decision
If RRT is expected to become necessary because of:
- Deteriorating clinical condition
- Progressive AKI course
- Increasing metabolic complications
- Significant pre-existing kidney disease
- Poor renal reserve
- Inability to maintain volume, electrolyte, or acid-base control conservatively
then it should be started without unnecessary delay.
However, in patients with non-life-threatening abnormalities or uncertainty about the need for RRT, the guideline supports:
- Conservative medical management
- Repeated clinical reassessment
- Monitoring of volume status, urine output, potassium, pH/bicarbonate, urea, creatinine, and clinical complications
2.3 What should not trigger RRT by itself
Do not start RRT solely because of:
- Isolated elevation of serum urea
- Isolated elevation of serum creatinine
- AKI stage alone, including KDIGO stage
- A goal of accelerating renal recovery
- A negative furosemide stress test alone
- Novel AKI biomarkers alone
Why?
Trials comparing early versus delayed initiation have generally not shown a clear mortality benefit from routine early RRT. Also, many patients assigned to delayed strategies recover without ever requiring RRT. Thus, indiscriminate early treatment risks exposing patients to catheter-related, bleeding, hemodynamic, and infectious complications without definite benefit.
Furosemide stress test
A negative test suggests a greater likelihood of AKI progression but does not itself establish an indication for dialysis. It is a prognostic aid, not a treatment trigger.
3. Diffusion, convection, and combined methods
Definitions
| Method | Main process | Main role |
|---|
| Diffusion | Solute movement down a concentration gradient | Efficient clearance of small solutes such as urea and potassium |
| Convection | Solute removal with ultrafiltrate flow, followed by replacement fluid | Removes small and middle molecules |
| Hemodiafiltration | Combination of diffusion and convection | Provides both mechanisms |
Guideline position
For ICU patients with AKI who need RRT:
- Diffusive, convective, and combined techniques produce similar patient outcomes.
- No method has shown superiority for mortality, renal recovery, or hemodynamic stability.
- The choice should depend on the treatment goal, local expertise, equipment, filter performance, and patient-specific circumstances.
Important practical points
Severe life-threatening hyperkalemia
Prefer a diffusive technique with high dialysate flow, if available.
Reason: Diffusion is more efficient for rapid removal of small molecules, especially potassium.
Sepsis
Diffusive, convective, or combined approaches can all be used. No modality has proven superior for survival in septic ICU patients requiring RRT.
Rhabdomyolysis
- Do not initiate RRT merely because creatine kinase or myoglobin is high.
- Start RRT only when there is AKI with a clear usual indication, such as refractory hyperkalemia, acidosis, fluid overload, or severe oliguria/anuria.
- Convective or high cut-off approaches may remove more myoglobin, but evidence has not consistently shown better patient outcomes.
Predilution versus postdilution in hemofiltration
- Either can be used.
- Predilution may be preferred when filters repeatedly clot.
- Predilution reduces hemoconcentration in the filter but requires a higher substitution volume to obtain similar solute clearance.
- Postdilution can give more efficient clearance but may increase hemoconcentration and clotting risk.
4. Continuous, intermittent, and prolonged intermittent RRT
Definitions
| Modality | Typical duration | Main characteristics |
|---|
| IHD: intermittent hemodialysis | 4-6 hours | Rapid solute and fluid shifts |
| PIRRT/SLED: prolonged intermittent RRT | 6-12 hours | Intermediate approach |
| CRRT: continuous RRT | ~24 hours/day | Gradual and continuous solute/fluid removal |
General recommendation
Continuous and intermittent modalities can be used equally in severe AKI with respect to survival. The modality should be selected according to the individual clinical situation.
When CRRT or prolonged RRT is preferred
Use continuous or prolonged modalities preferentially when there is:
- Hemodynamic instability
- High vasopressor requirement
- Recurrent intradialytic hypotension
- Need for slow, tightly controlled net ultrafiltration
- Concern about rapid osmotic shifts
- Increased intracranial pressure or cerebral edema risk
Rationale
CRRT provides slower and more stable correction of solute and fluid abnormalities. This reduces sudden changes in blood pressure, osmolality, urea concentration, and intravascular volume.
Increased intracranial pressure
RRT should be adapted to avoid dialysis disequilibrium and maintain appropriate serum osmolality.
Monitor regularly:
- Serum osmolality
- Sodium
- Urea
- Blood glucose
Rapid urea removal can lower extracellular osmolality, draw water into cells, and contribute to cerebral edema.
Thrombocytopenia
Intermittent and continuous therapies may both be considered. Thrombocytopenia alone does not dictate a specific modality.
Fluid overload
All modalities can achieve a negative fluid balance. Success depends more on individualized ultrafiltration targets and hemodynamic tolerance than on the modality itself.
Mobilization
CRRT should not delay early mobilization. Safe mobilization is possible even with CRRT and femoral access when proper protocols, trained staff, and monitoring are in place.
5. Anticoagulation during RRT
Why anticoagulation is used
Extracorporeal circuits activate coagulation and can clot. Anticoagulation helps preserve filter life and maintain prescribed treatment delivery.
Main options
- Regional citrate anticoagulation
- Systemic unfractionated heparin
- Systemic low-molecular-weight heparin
- Alternative agents for heparin-induced thrombocytopenia, such as argatroban
Citrate versus heparin
Regional citrate and systemic heparin are considered similar for major patient outcomes:
- Mortality
- Renal recovery
- Transfusion frequency
When citrate is preferred
Regional citrate anticoagulation is preferred when:
- Heparin is contraindicated
- The patient has active bleeding
- Bleeding risk is high
- Systemic heparin is associated with poor filter life or repeated clotting
Citrate in shock or liver failure
Shock and liver failure are not absolute contraindications to citrate anticoagulation.
However, citrate can only be used safely with regular monitoring of:
- Lactate
- Ionized calcium
- Total calcium
- Acid-base status
- Calcium replacement requirement
- Evidence of citrate accumulation
Avoid or use extreme caution
Avoid citrate where possible in patients with:
- Pronounced and progressive lactic acidosis due to shock
- Severe liver failure with impaired citrate metabolism
- Clinical or biochemical evidence of citrate accumulation
Citrate accumulation
If citrate accumulation occurs:
- Stop citrate anticoagulation
- Use RRT without anticoagulation or switch to heparin if appropriate
- Use bicarbonate-buffered replacement/dialysate fluid
Heparin
- LMWH and unfractionated heparin may both be used.
- Unfractionated heparin is often preferred in ICU practice because it:
- Has a shorter half-life
- Is easier to monitor
- Is easier to stop or reverse if bleeding occurs
Heparin-induced thrombocytopenia type II
- Stop all heparin products.
- Argatroban can be considered.
- It is hepatically metabolized and not significantly removed by standard dialysis filters.
- In critically ill patients, especially those with liver dysfunction, a much lower initial dose may be needed.
6. Dose of RRT
CRRT dose
The recommended delivered CRRT dose is:
20-25 mL/kg/hour
This usually refers to the delivered effluent dose.
Prescribed versus delivered dose
The actual delivered dose is often lower than the prescribed dose because of:
- Circuit clotting
- Machine downtime
- Procedures and transport
- Vascular access problems
- Filter changes
- Alarm-related interruptions
Therefore, a prescription around 25-30 mL/kg/hour may be required to reliably deliver 20-25 mL/kg/hour.
Intermittent dialysis dose
For IHD or prolonged intermittent therapy, dose should be individualized according to:
- Dialysis membrane
- Blood-flow rate
- Dialysate-flow rate
- Convective volume, if hemodiafiltration is used
- Session duration
- Frequency of sessions
- Potassium concentration
- Acid-base status
- Catabolic state
- Fluid balance
Dialysis disequilibrium syndrome
Particular attention is required when initiating intermittent treatment.
Mechanism
Rapid removal of urea reduces serum osmolality. Water then shifts into cells, especially brain cells, causing cerebral edema.
Possible manifestations
- Headache
- Nausea/vomiting
- Restlessness
- Confusion
- Seizures
- Reduced consciousness
Prevention
- Avoid overly rapid initial urea removal
- Adjust dialysis duration, flow rates, and solute clearance
- Consider slower or continuous approaches in high-risk patients
- Closely monitor neurological status and serum osmolality
High-volume hemofiltration
Do not routinely use high-volume hemofiltration for sepsis or septic shock.
- High-volume hemofiltration was defined as approximately 50 mL/kg/hour or more.
- Despite theories that it may remove inflammatory mediators, evidence is inconsistent and generally low quality.
- No reliable survival benefit has been established.
7. Pharmacotherapy and antimicrobial dosing during RRT
Core pharmacokinetic principle
Drug dosing in RRT must account for:
- Residual native kidney function
- RRT modality and treatment intensity
- Extracorporeal clearance
- Nonrenal clearance
- Volume of distribution
- Protein binding
- Severity of infection and likely pathogen
- Fluid accumulation, edema, and hypoalbuminemia
Initial or loading dose
The initial anti-infective dose should not routinely be reduced simply because the patient has AKI or is receiving RRT.
Why?
Critically ill patients often have an increased volume of distribution due to:
- Sepsis
- Capillary leak
- Fluid resuscitation
- Edema
- Hypoalbuminemia
This is especially important for hydrophilic drugs, such as:
- Beta-lactams
- Aminoglycosides
- Vancomycin
Underdosing the first dose can delay adequate antimicrobial exposure in life-threatening infection.
Administration strategy
Continuous RRT
For time-dependent antibiotics, continuous or prolonged administration under therapeutic drug monitoring may improve pharmacokinetic/pharmacodynamic target attainment.
Intermittent hemodialysis
For IHD, short or prolonged infusion may be preferred over continuous administration.
Post-IHD supplemental doses
After IHD, give an additional dose when indicated by the drug’s pharmacokinetics, the dialysis dose, and evidence-based drug-specific recommendations.
Therapeutic drug monitoring
Use TDM whenever available for critical antibiotics in RRT patients, particularly:
- Beta-lactams
- Vancomycin
- Aminoglycosides
TDM is especially important in patients with:
- Multiple organ failure
- Severe sepsis or septic shock
- Unpredictable residual kidney function
- High-intensity RRT
- Long or interrupted RRT sessions
- Marked fluid overload
- Risk of toxicity or treatment failure
The guideline notes that TDM improves attainment of pharmacokinetic/pharmacodynamic targets, although a consistent mortality benefit has not been demonstrated.
8. Stopping or weaning RRT
General principle
Discontinue RRT when the original absolute indication has resolved and native kidney function appears sufficient to maintain:
- Potassium control
- Acid-base balance
- Fluid balance
- Uremic solute control
- Clinical stability
Urine output
Urine output is a useful predictor of successful liberation from RRT.
A spontaneous urine output of approximately:
300-600 mL/day without diuretics
may suggest a reasonable possibility of successful discontinuation.
This is guidance rather than a universal threshold. The guideline cannot recommend one precise urine-output value for all patients.
Other factors suggesting weaning may fail
Persistent hyperkalemia
- Potassium >5.5 mmol/L may indicate a higher risk of failure.
- Investigate and correct causes before attempting weaning.
Persistent metabolic acidosis
- A pH <7.3 may indicate higher risk of failure.
- Identify and address the cause before discontinuing treatment.
Fluid overload
Correct significant fluid overload before stopping RRT where possible.
Kinetic GFR
Kinetic GFR based on endogenous clearance may be considered as an additional predictor of kidney recovery and concentrating capacity, but evidence remains limited.
Biomarkers
The guideline cannot recommend new kidney damage or functional biomarkers for predicting successful discontinuation.
Diuretics
Diuretics may be considered to increase urine output during discontinuation, but they should not be interpreted as proof of renal recovery. They do not replace assessment of solute clearance, potassium, acid-base status, and volume control.
Definition of successful discontinuation
The consensus considers cessation successful when RRT is not restarted for approximately 7 days after an intentional discontinuation attempt.
9. High-yield exam and clinical takeaways
- Start RRT immediately for life-threatening fluid, electrolyte, acid-base, or uremic complications.
- Do not dialyze solely because creatinine or urea is high.
- AKI stage, biomarkers, and a negative furosemide stress test should not independently determine RRT initiation.
- No clear survival advantage exists for diffusion versus convection.
- For severe hyperkalemia, use high-flow diffusive dialysis if available.
- CRRT, IHD, and PIRRT have similar survival outcomes overall.
- Prefer CRRT or PIRRT for hemodynamic instability or concern about rapid osmotic shifts.
- Use RRT carefully in increased intracranial pressure, with close osmolality and electrolyte monitoring.
- Citrate and heparin have similar major patient outcomes, but citrate is useful with bleeding risk or poor filter life on heparin.
- Shock or liver failure are relative, not absolute, contraindications to citrate.
- Delivered CRRT dose: 20-25 mL/kg/hour.
- High-volume hemofiltration is not recommended for sepsis or septic shock.
- Do not routinely reduce the antimicrobial loading dose in ICU patients on RRT.
- Use antibiotic TDM whenever feasible, especially for beta-lactams, vancomycin, and aminoglycosides.
- Urine output of 300-600 mL/day without diuretics may support a trial off RRT, but assess the entire patient.
- Persistent hyperkalemia, acidosis, or fluid overload predicts weaning failure.
Important caveat
These are detailed study notes, not a bedside protocol. Local ICU policy, modality availability, nephrology input, vascular access factors, and the patient’s hemodynamic and neurological condition must guide actual prescribing. The uploaded version was labeled an article-in-press manuscript, so the final published text should be checked for minor editorial changes.# Detailed Notes: Multidisciplinary Guidelines on Renal Replacement Therapy in Intensive Care Medicine
Source: Meersch-Dini et al., Critical Care (2025), DOI: 10.1186/s13054-025-05817-6.
Scope: Renal replacement therapy (RRT) for critically ill adults with acute kidney injury (AKI), particularly dialysis-dependent AKI in the ICU.
These notes summarize a guideline manuscript supplied by you. It was labelled “article in press,” so consult the final version and local ICU/nephrology protocol for clinical use.
1. Background and purpose
Why RRT matters in ICU
- AKI is common in critical illness.
- Severe AKI can produce life-threatening:
- Fluid overload and pulmonary edema
- Hyperkalemia
- Metabolic acidosis
- Uremic complications
- RRT supports kidney function while the underlying illness is treated. It does not by itself reverse the cause of AKI.
Aim of the guideline
To update guidance on the use of RRT in ICU patients using current evidence and expert consensus.
Development process
- Binational multidisciplinary panel from Germany and Austria.
- Included intensivists, anesthesiologists, nephrologists, internists, surgeons, other professionals, and patient representatives.
- Evidence searched in PubMed, Scopus, and Cochrane databases.
- Recommendations developed using:
- PICO questions
- Systematic literature review
- Modified Delphi process
- GRADE approach
- Multiple full-panel consensus conferences
Seven areas addressed
- Starting RRT
- Diffusion versus convection
- Continuous versus intermittent therapy
- Anticoagulation
- RRT dose
- Pharmacotherapy during RRT
- Stopping RRT
2. Starting RRT
Core principle
Start RRT for clinical indications, not just because creatinine or urea is high.
Absolute, urgent indications
RRT should be started immediately for life-threatening disturbance of:
| Problem | Clinical examples |
|---|
| Fluid balance | Refractory fluid overload, pulmonary edema |
| Electrolytes | Severe or refractory hyperkalemia |
| Acid-base status | Severe metabolic acidosis not responsive to medical treatment |
| Uremia | Uremic complications such as encephalopathy, pericarditis, severe symptomatic uremia |
Recommendation 1.1
Start RRT immediately in life-threatening fluid, acid-base, or electrolyte disturbances.
Relative indications and timing
If RRT is likely to become necessary because of the patient’s illness trajectory, underlying disease, or worsening AKI, it should not be unnecessarily delayed.
However, if there is no immediate threat and the need for RRT is uncertain:
- Optimize conservative therapy.
- Treat reversible causes.
- Reassess frequently.
- Avoid exposing patients to unnecessary catheterization and extracorporeal treatment.
Recommendation 1.2
If RRT is expected to become necessary based on clinical course and comorbidities, initiate it without further delay.
Recommendation 1.3
For non-life-threatening abnormalities or uncertain need for RRT, use conservative measures and reassess regularly.
What should not trigger RRT alone?
Isolated urea or creatinine elevation
- A biochemical elevation alone, without symptoms or complications of kidney failure, is not a sufficient reason to initiate RRT.
- The decision must integrate:
- Clinical examination
- Fluid status
- Potassium
- Acid-base status
- Urine output
- Uremic symptoms
- Overall trajectory and prognosis
Recommendation 1.5
RRT may be deferred in isolated elevation of serum urea or creatinine without clinical signs attributable to kidney failure.
Furosemide stress test
- A poor diuretic response can predict AKI progression.
- It does not by itself establish a need for RRT.
Biomarkers
- Novel biomarkers may help risk-stratify AKI.
- Current evidence does not support using biomarkers alone to decide when to start RRT.
Recommendation 1.7
Do not use biomarkers alone to make the RRT initiation decision.
Evidence on early versus delayed initiation
- Trials and systematic reviews have not shown a consistent mortality benefit from routine early RRT before conventional clinical indications occur.
- A major limitation is that many patients assigned to delayed initiation improve and never need RRT.
- Therefore, indiscriminate early RRT may expose some patients to:
- Vascular access complications
- Bleeding
- Infection
- Hypotension
- Unnecessary treatment burden
Exam takeaway
Do not start RRT solely for KDIGO stage, creatinine, BUN, a failed furosemide stress test, or a biomarker result. Start for dangerous complications or a strongly anticipated clinical need.
3. Diffusion versus convection
Definitions
Diffusion
- Solute transport occurs down a concentration gradient across a semipermeable membrane.
- Particularly effective for small molecules:
- Urea
- Creatinine
- Potassium
- Examples:
- Intermittent hemodialysis (IHD)
- Continuous venovenous hemodialysis (CVVHD)
Convection
- Solutes are dragged across the membrane with water movement, called solvent drag.
- Used in hemofiltration and hemodiafiltration.
- Examples:
- Continuous venovenous hemofiltration (CVVH)
- Continuous venovenous hemodiafiltration (CVVHDF)
Main guidance
Diffusive, convective, and combined techniques have broadly similar patient outcomes in ICU AKI.
Recommendation 2.1
In AKI requiring RRT, diffusion, convection, or a combined method can be used because outcomes are similar.
Sepsis
For patients with sepsis requiring RRT:
- No modality has proven superior for mortality or renal recovery.
- Diffusion, convection, and combined methods can all be used.
Recommendation 2.3
Use diffusive, convective, or combined techniques equally in septic ICU patients needing RRT.
Severe hyperkalemia
- Potassium is a small molecule and is cleared efficiently by diffusion.
- In life-threatening hyperkalemia, if available, favor a diffusive technique with high dialysate flow.
Recommendation 2.4
For severe life-threatening hyperkalemia, prefer high-dialysate-flow diffusive RRT over a convective method.
Predilution versus postdilution in hemofiltration
| Feature | Predilution | Postdilution |
|---|
| Replacement fluid | Given before filter | Given after filter |
| Hemoconcentration in filter | Less | More |
| Filter clotting risk | Lower | Higher |
| Solute clearance efficiency | Lower for same volume | Higher |
| Best consideration | Repeated filter clotting | When maximizing clearance is important |
Recommendation 2.2
Either pre- or postdilution may be used. Prefer predilution when filter clotting is recurrent.
Rhabdomyolysis
- Myoglobin contributes to tubular injury and AKI.
- Convective methods or high cut-off membranes may increase myoglobin removal.
- But improved myoglobin clearance has not consistently improved patient outcomes.
- Do not begin RRT solely because myoglobin is high.
Recommendation 2.5
Initiate RRT in rhabdomyolysis only when AKI has a clear RRT indication.
4. Continuous versus intermittent RRT
Modalities
| Modality | Typical treatment pattern | Main characteristic |
|---|
| IHD | 4-6 hours/session | Rapid solute and fluid removal |
| PIRRT/SLED | About 6-12 hours | Slower, prolonged intermittent therapy |
| CRRT | Continuous, usually 24 hours/day | Gradual, sustained solute and fluid removal |
Mortality and renal recovery
- No clear survival advantage of CRRT over intermittent RRT in unselected ICU patients with severe AKI.
- Renal recovery is also broadly comparable.
- Select a modality according to the patient’s hemodynamics, neurological status, fluid requirements, logistics, and staff expertise.
Recommendation 3.1
Continuous and intermittent RRT can be used equally for survival in severe AKI.
Recommendation 3.2
Individualize the modality to the clinical situation.
Hemodynamic instability
CRRT or prolonged intermittent therapy is generally preferred in patients who are:
- Hypotensive
- Receiving vasopressors
- Unable to tolerate rapid fluid removal
- At risk of intradialytic hypotension
Reason:
- Slower fluid and solute shifts
- Better hemodynamic tolerance
Recommendation 3.3
Prefer continuous or prolonged RRT in hemodynamically unstable patients to promote stability and reduce hypotension.
Raised intracranial pressure
Rapid decline in serum urea/osmolality can worsen cerebral edema and precipitate dialysis disequilibrium.
Recommendation 3.5
Increased intracranial pressure requires RRT adapted to preserve appropriate serum osmolality.
Monitor:
- Serum osmolality
- Sodium
- Urea
- Blood glucose
Avoid:
- Abrupt urea removal
- Large rapid osmotic shifts
- Aggressive fluid shifts
Fluid overload
- IHD, PIRRT, and CRRT can all create a negative fluid balance.
- The key determinant is not the modality alone but a patient-specific ultrafiltration plan.
- CRRT may facilitate steady fluid removal in unstable patients.
Recommendation 3.6
Any modality can be used for negative fluid balance. Tailor ultrafiltration to the individual patient.
Mobilization
- CRRT should not automatically delay early mobilization.
- Safe mobilization is feasible during CRRT with appropriately trained staff and line-safety procedures.
Recommendation 3.7
Do not delay or prevent early mobilization merely because the patient is receiving CRRT.
5. Anticoagulation in RRT
Goals
Anticoagulation should:
- Maintain filter patency
- Reduce circuit clotting
- Minimize interruptions
- Avoid bleeding and metabolic complications
Main options
- Regional citrate anticoagulation (RCA)
- Systemic unfractionated heparin (UFH)
- Low-molecular-weight heparin (LMWH)
- Alternatives in heparin-induced thrombocytopenia, such as argatroban
Citrate versus heparin
- Citrate and systemic heparin have similar major patient outcomes, including mortality and renal recovery.
- Citrate is useful when systemic anticoagulation is undesirable, particularly with bleeding risk.
Recommendation 4.1
Regional citrate and systemic heparin can be used equally regarding major patient outcomes.
Recommendation 4.2
Prefer regional citrate in patients with:
- Heparin contraindication
- Active bleeding
- High risk of bleeding
Recommendation 4.3
If systemic heparin results in short filter life and inadequate delivery of therapy, consider switching to citrate.
Citrate in shock or liver failure
Older practice often treated liver failure and shock as contraindications to citrate. The guideline states they are not absolute contraindications.
However, citrate should only be used with close metabolic surveillance.
Monitor during RCA
- Ionized calcium
- Total calcium
- Total calcium to ionized calcium relationship
- Lactate
- Acid-base balance
- Evidence of citrate accumulation
Recommendation 4.4
Citrate can be used in shock or liver failure with regular lactate, ionized calcium, and total calcium monitoring.
Avoid or use extreme caution
Avoid citrate where possible in:
- Pronounced progressive lactic acidosis
- Shock with worsening lactate
- Severe liver failure with impaired citrate metabolism
Recommendation 4.6
Avoid RCA if possible in progressive lactic acidosis related to shock and severe liver failure.
Citrate accumulation
Possible clues:
- Increasing total calcium requirement
- Worsening hypocalcemia despite calcium infusion
- Rising total calcium relative to ionized calcium
- Metabolic derangement
- Rising lactate in the appropriate clinical setting
Recommendation 4.5
If citrate accumulation occurs:
- Stop citrate.
- Switch to no anticoagulation or heparin-based CRRT.
- Use bicarbonate as buffering agent.
Heparin
- UFH and LMWH may both be considered.
- UFH is often favored in ICU because:
- Easier monitoring
- Shorter half-life
- More readily reversible
Recommendation 4.7
UFH and LMWH can both be used, but UFH is preferred when rapid adjustment and monitoring are needed.
Heparin-induced thrombocytopenia type II
- Stop all heparin exposure.
- Argatroban is a possible alternative.
- It is hepatically metabolized, so dose reduction is needed in liver dysfunction.
Recommendation 4.8
Argatroban may be considered in acute HIT type II.
6. RRT dose
Continuous RRT dose
Recommendation 5.1
Deliver CRRT at 20-25 mL/kg/hour.
Important distinction
- Prescribed dose is the dose ordered.
- Delivered dose is what the patient actually receives.
Interruptions reduce delivered dose:
- Filter clotting
- Circuit changes
- Procedures and transport
- Imaging
- Access problems
- Machine downtime
Therefore, a slightly higher prescribed dose, such as 25-30 mL/kg/hour, may be required to achieve a delivered dose of 20-25 mL/kg/hour.
Intermittent RRT dose
Dose is individualized according to:
- Dialysis membrane
- Blood flow
- Dialysate flow
- Convective volume in hemodiafiltration
- Duration of treatment
- Frequency of sessions
- Potassium concentration
- Acidosis severity
- Catabolic state
- Fluid needs
Recommendation 5.2
Determine intermittent-dialysis dose using treatment parameters and clinical/laboratory targets, especially electrolyte and acid-base control.
Dialysis disequilibrium syndrome
Mechanism
- Rapid urea removal lowers extracellular osmolality.
- Water shifts into cells, especially brain cells.
- Cerebral edema can occur.
Possible manifestations
- Headache
- Nausea/vomiting
- Restlessness
- Confusion
- Seizures
- Reduced consciousness
Higher-risk situations
- Very high urea
- First dialysis session
- Rapidly delivered IHD
- Acute neurological injury or raised intracranial pressure
Recommendation 5.3
When initiating RRT, assess risk of disequilibrium and adjust dose accordingly.
Practical approach:
- Use gentler initial clearance.
- Reduce blood and/or dialysate flow as appropriate.
- Shorten first session.
- Consider prolonged or continuous therapy where indicated.
- Monitor neurological status and osmolality-related variables.
High-volume hemofiltration
- Defined in this guideline as approximately 50 mL/kg/hour or more.
- Proposed to remove inflammatory mediators in sepsis.
- Evidence does not show a reliable survival benefit.
- It increases complexity and may increase unintended removal of useful solutes or drugs.
Recommendation 5.4
Do not use high-volume hemofiltration routinely in sepsis or septic shock.
7. Pharmacotherapy during RRT
Why antibiotic dosing is difficult
Drug exposure in ICU patients receiving RRT is highly variable because of:
- Residual renal function
- Nonrenal clearance
- RRT modality
- Dialyzer membrane
- Blood and dialysate flow
- Effluent dose
- Drug protein binding
- Molecular size
- Volume of distribution
- Capillary leak, edema, hypoalbuminemia
- Circuit downtime and interruptions
Loading dose
Principle
The initial dose depends mainly on the volume of distribution, not on renal clearance.
In severe critical illness:
- Edema, capillary leak, and hypoalbuminemia commonly increase volume of distribution.
- Hydrophilic drugs may distribute into expanded extracellular fluid, resulting in low initial plasma concentration.
Hydrophilic antibiotics include:
- Beta-lactams
- Aminoglycosides
- Vancomycin
Recommendations 6.1-6.3
- Base starting dose on the current volume of distribution.
- Do not routinely reduce the initial anti-infective dose because the patient is receiving RRT.
- Give the loading dose as a short infusion to achieve rapid effective concentration, even if continuous or prolonged infusion will follow.
Maintenance dose
Principle
Maintenance dosing depends on total clearance:
Total clearance = residual kidney clearance + extracorporeal/RRT clearance + nonrenal clearance
Recommendation 6.4
If effective levels are not known, calculate maintenance dose using estimated total clearance, including residual renal function and machine clearance.
Infusion strategy
Time-dependent antibiotics
Examples include many beta-lactams.
- In CRRT, prolonged or continuous infusion may provide more reliable pharmacokinetic/pharmacodynamic target attainment, particularly if TDM is available.
- In IHD, continuous infusion may be less suitable because dialysis interruption changes drug clearance. Short or prolonged infusion can be preferred.
Recommendations
- 6.5: Consider prolonged or continuous administration of time-dependent antibiotics in continuous RRT, ideally with TDM.
- 6.6: In IHD, short or prolonged infusion may be preferable to continuous infusion.
- 6.7: Following IHD, give a supplemental dose when required to restore therapeutic concentration, according to the drug’s pharmacokinetics and accepted dosing guidance.
Therapeutic drug monitoring
Recommendation 6.8
Use TDM when available for high-risk antibiotics in RRT, including:
- Beta-lactams
- Vancomycin
- Aminoglycosides
The guideline also highlights evidence supporting TDM consideration for:
- Linezolid
- Colistin
- Possibly daptomycin, tigecycline, and fluoroquinolones
Why TDM is important
TDM improves the probability of achieving pharmacokinetic/pharmacodynamic targets and can reduce nephrotoxicity, though a consistent mortality benefit has not been proven.
8. Stopping RRT and renal recovery
Principle
Stop RRT when the original indication has resolved and native kidney function is sufficient to maintain:
- Potassium homeostasis
- Acid-base balance
- Fluid balance
- Solute control
A successful discontinuation was considered as no restart of RRT within 7 days.
Urine output
Recommendation 7.2
Use urine output as a predictor of successful RRT discontinuation.
Recommendation 7.3
There is no exact mandatory urine output threshold. As practical guidance, spontaneous urine output of 300-600 mL/day without diuretics may suggest a reasonable chance of successful discontinuation.
Important:
- Urine output predicts recovery but is not sufficient alone.
- A patient may produce urine but still have inadequate clearance, persistent hyperkalemia, acidosis, or fluid overload.
Diuretics
Recommendation 7.1
Diuretics may be considered to increase urine output during RRT discontinuation.
But:
- They do not prove renal recovery.
- They should not substitute for reassessing clearance and complications.
- They should not be used to force a trial off RRT in a patient with unresolved indications.
Other predictors and barriers
Kinetic GFR
Kinetic GFR based on endogenous clearance may be considered as an adjunctive predictor after stopping RRT.
Persistent hyperkalemia
A potassium concentration above 5.5 mmol/L suggests a higher risk of failed RRT withdrawal.
Recommendation 7.5
Investigate persistent hyperkalemia before attempting weaning.
Persistent metabolic acidosis
A pH below 7.3 may predict failure of discontinuation.
Recommendation 7.7
Treat and investigate persistent metabolic acidosis before stopping RRT.
Fluid overload
Fluid overload should be corrected before attempting discontinuation.
Recommendation 7.8
Correct fluid overload before trialing RRT withdrawal.
Biomarkers
Current evidence is inadequate to recommend novel kidney injury or functional biomarkers for stopping RRT.
9. Practical ICU decision framework
A. Should RRT be started?
- Assess for emergency indications:
- Refractory hyperkalemia
- Severe acidosis
- Pulmonary edema/fluid overload
- Uremic complication
- If none:
- Treat reversible factors.
- Review fluid status, urine output, hemodynamics, trajectory, and comorbidity.
- Reassess regularly.
- Do not initiate solely for:
- Creatinine
- Urea
- AKI stage
- Failed furosemide stress test
- Biomarker elevation
B. Which modality?
- Hemodynamic instability: CRRT or PIRRT often favored.
- Severe hyperkalemia needing rapid clearance: high-flow diffusive therapy if available.
- Raised intracranial pressure: avoid rapid osmotic changes, use carefully adapted therapy.
- Fluid removal: any modality can work, but tailor ultrafiltration rate.
- Stable patient: IHD, PIRRT, or CRRT based on clinical goals and local capability.
C. Anticoagulation choice
- Standard options: citrate or systemic heparin.
- Bleeding risk or heparin contraindication: favor citrate if feasible.
- Shock/liver failure: citrate may still be possible with close calcium and lactate monitoring.
- HIT II: avoid heparin, consider argatroban.
D. Dose
- CRRT delivered dose: 20-25 mL/kg/hour.
- Anticipate downtime and prescribe enough to achieve this.
- Avoid routine high-volume hemofiltration in septic shock.
- Prevent dialysis disequilibrium in high-risk patients.
E. Antibiotics
- Do not reduce loading dose merely because RRT is used.
- Individualize maintenance dose according to total clearance.
- Use TDM when available for high-risk drugs.
F. When to stop
- Original indication resolved.
- Adequate urine output and recovery trend.
- No persistent hyperkalemia, acidosis, or fluid overload.
- Monitor closely for at least the next week for restart requirement.
10. High-yield exam points
- Emergency RRT indications: refractory hyperkalemia, severe acidosis, refractory fluid overload/pulmonary edema, and symptomatic uremia.
- Creatinine alone is not an indication for RRT.
- Routine “early” RRT before clinical indications does not have a consistent mortality benefit.
- Diffusion and convection have similar overall outcomes in ICU AKI.
- For life-threatening hyperkalemia, use a high-flow diffusive technique if possible.
- CRRT has no established overall mortality superiority, but is often preferred in hemodynamic instability.
- In raised intracranial pressure, avoid rapid osmotic shifts and monitor osmolality, sodium, urea, and glucose.
- Citrate is preferred with bleeding risk or heparin contraindication, but requires calcium and metabolic monitoring.
- Shock and liver failure are not absolute contraindications to citrate anticoagulation.
- Target delivered CRRT dose: 20-25 mL/kg/hour.
- Routine high-volume hemofiltration is not recommended for sepsis/septic shock.
- Do not reduce the initial antibiotic loading dose merely due to RRT.
- TDM is particularly useful for beta-lactams, vancomycin, and aminoglycosides.
- Urine output helps predict successful liberation from RRT. A spontaneous output of 300-600 mL/day is a useful guide, not an absolute rule.
- Persistent potassium >5.5 mmol/L, pH <7.3, or fluid overload should prompt caution before stopping RRT.
The guideline is available in the open-access
full guideline article. A recent systematic review also suggests that lower CRRT intensity may be comparable to standard intensity, but this is post-guideline evidence and should be interpreted alongside local protocols (
2026 review, PMID 41896891).