Dialysis in the ICU is a temporary form of blood filtration used when a critically ill patient’s kidneys can no longer clear waste, balance fluids, or regulate blood chemistry on their own. The formal term is kidney replacement therapy, or KRT, and roughly a third of ICU patients with acute kidney injury end up needing it. Unlike outpatient dialysis for chronic kidney disease, ICU dialysis is usually a crisis intervention, run at the bedside around the clock or in extended sessions, tailored to patients who are too unstable for a standard dialysis unit. The reasons it is needed, the way it is delivered, and the decisions that surround it are more complex than most people realize.
Why ICU Patients Need Kidney Support
Healthy kidneys filter blood continuously, removing toxins, balancing electrolytes, and pulling off excess fluid. When a serious illness causes those kidneys to fail acutely, the consequences compound fast. Potassium can spike to levels that trigger dangerous heart rhythms. Acid builds up in the blood. Fluid accumulates in the lungs, making breathing harder. Waste products like urea climb to toxic concentrations. In the ICU, the most common trigger for this kind of acute kidney injury is not a problem with the kidneys themselves. It is a problem with the rest of the body. In one study of 100 ICU patients with acute kidney injury, the vast majority of cases stemmed from reduced blood flow to the kidneys rather than direct kidney damage, with fluid loss accounting for nearly half of cases and septic shock, drug toxicity, and heart failure making up most of the remainder.1European Journal of Cardiovascular Medicine. To Study the Risk Factors, Treatment and Outcomes of Acute Kidney Injury in Intensive Care Unit Patients
Sepsis is an especially powerful driver. A large prospective study found that sepsis was the precipitating cause in over 70% of cases of severe acute kidney injury requiring dialysis, and community-acquired pneumonia was the single most common admission diagnosis.2PLOS ONE. A real-world prospective study on dialysis-requiring acute kidney injury In sepsis, the body’s inflammatory response causes widespread drops in blood pressure and disrupts the fine blood vessels inside the kidneys. Even with aggressive fluid resuscitation and vasopressor drugs, the kidneys can shut down. When they do, dialysis becomes the only way to keep the internal environment stable enough for the rest of the body to have a chance at recovery.
The Three Main Approaches
ICU dialysis is not one-size-fits-all. Critically ill patients vary enormously in how stable their blood pressure is, how much fluid needs to come off, and how long their kidneys are expected to be offline. Three broad approaches exist, each with trade-offs.
Continuous kidney replacement therapy, or CKRT, runs around the clock, typically for 24 hours or more at a time. Because it works slowly and steadily, it pulls fluid and waste at a gentle rate that is easier on blood pressure. That makes it the go-to choice for patients who are hemodynamically unstable, meaning their circulation is already fragile from shock, heart failure, or large doses of blood-pressure-supporting drugs.3American Journal of Kidney Diseases. Continuous Kidney Replacement Therapies: Core Curriculum 2025 The downside is that continuous therapy ties up a specialized machine and demands constant nursing attention, and patients must remain anticoagulated (their blood kept from clotting in the circuit) for extended periods.
Intermittent hemodialysis, or IHD, is essentially the same technology used in outpatient dialysis centers but adapted for the ICU. A session typically lasts three to five hours. It clears waste and fluid rapidly, which is useful in emergencies like severe potassium elevations. It is more practical in terms of cost, does not always require anticoagulation, and frees the patient from being connected to a machine all day.4PubMed Central. Single-best Choice Between Intermittent Versus Continuous Renal Replacement Therapy: A Review But the fast fluid removal can cause blood pressure to drop in already-unstable patients, which limits its use in the sickest cases.
A third option, sustained low-efficiency dialysis (SLED), splits the difference. It uses a standard dialysis machine but runs for six to twelve hours at lower flow rates, providing gentler fluid and solute removal than standard IHD while avoiding the need for a dedicated continuous-therapy machine. Studies comparing SLED with continuous therapy have found similar mortality rates and hemodynamic stability, but SLED has been associated with fewer blood transfusions, shorter ventilator time, and lower costs.5PubMed Central. Sustained low-efficiency dialysis—an old method but possibly a new solution for environmental nephrology In practice, the choice among these modalities depends on what equipment is available, how stable the patient is, and the expertise of the ICU team.
Does It Matter Which Type You Get?
This is one of those questions where the research has been frustratingly inconclusive. Intuitively, continuous therapy seems like it should be better for the sickest patients because it avoids the sharp swings of intermittent treatment. But a secondary analysis of two large French randomized trials found that, among the most severely ill patients, there was no detectable survival difference between continuous therapy and intermittent hemodialysis. Among less severely ill patients, IHD was actually associated with better 60-day survival.6PubMed Central. Continuous renal replacement therapy versus intermittent hemodialysis as first modality for renal replacement therapy in severe acute kidney injury: a secondary analysis of AKIKI and IDEAL-ICU studies That finding runs counter to the assumption that continuous is always safer. In reality, the strongest factor in outcomes is the overall severity of the patient’s illness, not the brand of dialysis running at the bedside.
Getting Vascular Access
Before any form of ICU dialysis can begin, the patient needs a large-bore catheter placed into a major vein, usually in the neck (internal jugular), the groin (femoral), or occasionally below the collarbone (subclavian). These catheters have two channels, one to pull blood out to the machine and one to return it. The procedure is done at the bedside under ultrasound guidance, which has been shown to reduce mechanical complications like puncturing an artery or collapsing a lung.7Kidney International. Temporary hemodialysis catheters: recent advances
Where the catheter goes matters. A landmark trial found that femoral catheters were faster and easier to place than jugular ones when relying on anatomical landmarks, and rates of catheter-tip colonization by bacteria were not significantly different between the two sites.8PubMed Central. Vascular access sites for acute renal replacement in intensive care units Still, the primary catheter-related complication is clotting. A symptomatic deep vein thrombosis occurs in a small fraction of patients, but asymptomatic clots can be found in roughly 10 to 22% of cases when ultrasound is used to look for them.9Anaesthesia Critical Care & Pain Medicine. Haemodialysis catheters in the intensive care unit Infection is the other major risk, and implementing standardized sterile-insertion protocols has significantly reduced bloodstream infections associated with these catheters.
Keeping the Blood From Clotting in the Circuit
Any time blood travels through plastic tubing and a filter, it wants to clot. Clotting fouls the filter, interrupts treatment, and wastes resources. To prevent this, ICU teams use anticoagulation. The two main options are heparin, a classic blood thinner given systemically or regionally, and citrate, which is infused into the circuit to bind calcium (a necessary ingredient for clotting) and then reversed before the blood returns to the patient.
Citrate has emerged as the preferred approach for continuous therapy. A meta-analysis of randomized controlled trials found that citrate reduced the risk of circuit clotting compared to both regional and systemic heparin, and cut bleeding complications by roughly two-thirds compared to systemic heparin.10PubMed. Citrate versus heparin anticoagulation for continuous renal replacement therapy: an updated meta-analysis of RCTs A large randomized trial confirmed these findings: filters lasted a median of 47 hours with citrate versus 26 hours with heparin, and bleeding events were about a third as common.11JAMA. Effect of Regional Citrate Anticoagulation vs Systemic Heparin Anticoagulation During Continuous Kidney Replacement Therapy on Dialysis Filter Life Span and Mortality Among Critically Ill Patients With Acute Kidney Injury However, that same trial also found more new infections in the citrate group, a finding that remains unexplained. Neither approach produced a significant difference in mortality. Citrate does come with its own metabolic quirks, particularly the risk of low calcium levels, which requires careful monitoring.
How Much Dialysis Is Enough
Early on, there was hope that giving patients higher “doses” of dialysis, meaning faster flow rates through the filter, would lead to better outcomes. Two large multicenter trials put this to the test and found that effluent flow rates above 25 milliliters per kilogram of body weight per hour did not improve survival.12PubMed Central. Clinical review: Optimal dose of continuous renal replacement therapy in acute kidney injury Current guidelines recommend prescribing a dose in the range of 20 to 25 mL/kg/hr for continuous therapy.13PubMed Central. Continuous Renal Replacement Therapy Dosing in Critically Ill Patients: A Quality Improvement Initiative Even so, the dose a patient actually receives often falls short of what is prescribed, because machine downtime for filter changes, imaging studies, or procedures eats into treatment hours. One quality-improvement initiative found that only about two-thirds of treatments consistently delivered the recommended dose. This gap between prescribed and delivered therapy is considered a meaningful quality indicator worth tracking.14PubMed Central. Dose of Continuous Renal Replacement Therapy in Critically Ill Patients: A Bona Fide Quality Indicator
The Timing Debate
One of the most contested questions in critical care nephrology is when to start dialysis for a patient whose kidneys are failing but who does not yet have a life-threatening emergency like a dangerously high potassium level or severe fluid overload. The intuition that starting earlier should be better has not held up. The STARRT-AKI trial, one of the largest to address the question, enrolled nearly 3,000 critically ill patients and randomized them to either an accelerated start or a standard (watchful-waiting) strategy. Mortality at 90 days was virtually identical between the two groups. But patients in the accelerated group had more adverse events, and among survivors, more of them remained dependent on dialysis at 90 days compared to those who started later.15PubMed. Timing of Initiation of Renal-Replacement Therapy in Acute Kidney Injury
A meta-analysis pooling multiple trials came to a similar conclusion: early initiation did not reduce 28-day or overall mortality, and was associated with more episodes of low blood pressure and infection.16PubMed Central. The impact of early versus late initiation of renal replacement therapy in critically ill patients with acute kidney injury on mortality and clinical outcomes: a meta-analysis A key detail from the STARRT-AKI trial underscores why a watchful-waiting approach has value: nearly 40% of patients assigned to the standard strategy never needed dialysis at all, because their kidneys recovered on their own. Starting everyone early means exposing some patients to an invasive treatment they never actually required.
There may be exceptions. A retrospective analysis found that in patients with less severe organ failure, each hour of delay in starting continuous therapy was associated with a small but measurable increase in 28-day mortality and fewer days free of the ICU and ventilator.17Scientific Reports. The impact of initiation timing of continuous renal replacement therapy on outcomes in critically ill patients with acute kidney injury a retrospective study from the MIMIC-IV database This suggests that the right answer may depend on how sick the patient is overall, not on a single lab value or a fixed timeline.
Fluid Overload as a Target
Beyond clearing toxins, one of the most important functions of ICU dialysis is removing excess fluid. Critically ill patients receive enormous volumes of intravenous fluids during resuscitation, and when the kidneys cannot keep up, that fluid accumulates. It pools in the lungs, worsens breathing, swells tissues, and raises pressure inside the abdomen. Fluid overload exceeding 10% of body weight has been independently associated with an 82% increased odds of hospital death and fewer days off the ventilator, even after accounting for the severity of illness and the timing of dialysis initiation.18Critical Care Medicine. Fluid Overload Associates With Major Adverse Kidney Events in Critically Ill Patients With Acute Kidney Injury Requiring Continuous Renal Replacement Therapy This makes managing fluid balance a central goal of treatment, and in many cases it is the fluid removal, not the waste clearance, that drives the decision to start dialysis.
Deciding When to Stop
If starting dialysis is controversial, stopping it is arguably harder. When a patient’s kidneys begin recovering, the team wants to know whether they can safely disconnect the machine without the kidneys failing again. The most studied predictor is urine output: if a patient starts making more urine while on dialysis, it suggests the kidneys are waking up. A systematic review found that urine output before stopping dialysis had moderate accuracy for predicting successful liberation, with a sensitivity of about 66% and a specificity of roughly 74%.19PubMed Central. Determining the optimal time for liberation from renal replacement therapy in critically ill patients: a systematic review and meta-analysis (DOnE RRT) That is better than a coin flip but far from definitive. No optimal threshold for urine output could be determined, because studies used different cutoffs. Researchers are working on machine-learning models that integrate trends in vital signs, lab values, and fluid balance to improve these predictions.20PubMed Central. Predictive approach for liberation from acute dialysis in ICU patients using interpretable machine learning
What Happens After You Leave the ICU
Surviving ICU dialysis does not necessarily mean returning to normal kidney function. A 10-year prospective study found that among survivors, about 56% had complete kidney recovery by the time they left the hospital. None of those patients developed chronic kidney disease during follow-up. But 90% of survivors who left the hospital with only partial kidney recovery went on to have ongoing chronic kidney disease, and about 5% of all survivors eventually progressed to permanent kidney failure requiring lifelong dialysis.21PubMed Central. Long-term outcomes of survivors of ICU acute kidney injury requiring renal replacement therapy: a 10-year prospective cohort study
A nationwide cohort study put the five-year risk of end-stage kidney disease at about 3.8% for patients who survived dialysis-requiring acute kidney injury, compared to just 0.3% for other ICU patients. After adjusting for other factors, surviving ICU dialysis was associated with roughly a six-fold higher risk of permanent kidney failure over the following five years.22PubMed Central. Five-year risk of end-stage renal disease among intensive care patients surviving dialysis-requiring acute kidney injury: a nationwide cohort study This means that patients who needed ICU dialysis benefit from long-term follow-up with a kidney specialist, even if their function looked acceptable at discharge.
Nutrition During ICU Dialysis
Dialysis does not just remove toxins and excess fluid. It also strips out amino acids, water-soluble vitamins, and trace minerals. In a patient who is already critically ill, often unable to eat, and burning through calories at an accelerated rate, these losses compound the challenge of providing adequate nutrition. Kidney dysfunction can affect up to half of all ICU patients, making nutritional planning around dialysis a significant clinical concern.23PubMed Central. Nutrition support for patients with renal dysfunction in the intensive care unit: A narrative review Protein targets tend to be higher for patients on continuous therapy than for other ICU patients, because the dialysis filter continuously removes amino acids. Getting this balance wrong in either direction, too little protein and the patient wastes away, too much and waste products accumulate faster, requires ongoing adjustment by the clinical team.
The Nursing and Cost Reality
Running dialysis in the ICU is labor-intensive. Continuous therapy demands that a bedside nurse monitor the machine, manage anticoagulation, track fluid balance, respond to alarms, and change filters when they clot. A workload analysis found that continuous therapy consumed about 12 to 13% of a nurse’s shift, with the effort split between preparing and maintaining the circuit, drawing labs, and adjusting settings.24Intensive and Critical Care Nursing. Continuous venovenous renal replacement therapy in critically ill patients: A work load analysis That may not sound like much, but these patients simultaneously need ventilator management, medications, wound care, and repositioning. A multinational study found wide variation in nursing costs across regions, with continuous therapy generally more expensive in consumables and intermittent therapy more expensive in nursing time in most settings.25PubMed Central. Cost of acute renal replacement therapy in the intensive care unit: results from The Beginning and Ending Supportive Therapy for the Kidney (BEST Kidney) Study These practical realities influence which modality a hospital can offer. Many facilities in resource-limited settings rely on intermittent hemodialysis or SLED because they lack the equipment, staffing, or supply chain for continuous therapy.
When Dialysis Becomes an Ethical Question
In some ICU patients, the question is not how to optimize dialysis but whether to offer it at all. When a patient has advanced, irreversible multi-organ failure and no realistic prospect of recovery, dialysis can become a form of life prolongation without meaningful benefit. Clinical practice guidelines encourage shared decision-making between the medical team, the patient (when possible), and the family, with time-limited trials of dialysis as a useful tool. A time-limited trial means starting dialysis with a pre-set reassessment point, giving the kidneys and the rest of the body a defined window to respond before the team and family revisit goals of care.26PubMed Central. Withholding and withdrawing dialysis in the intensive care unit: benefits derived from consulting the renal physicians association/american society of nephrology clinical practice guideline, shared decision-making in the appropriate initiation of and withdrawal from dialysis
Surveys of ICU physicians reveal the complexity of these decisions. Most clinicians believe they have adequate criteria for defining when dialysis is futile, but many admit they would start or continue dialysis at a family’s request even when they judged it to be futile. Withholding dialysis (never starting it) and withdrawing it (stopping it once started) are perceived differently by many clinicians, even though ethically they are considered equivalent. Psychological barriers make withdrawal harder in practice, and many teams report that ethics committees are a valuable resource for navigating these situations.27PubMed. Bioethical issues related to continuous renal replacement therapy in intensive care patients
Dialysis for the Smallest Patients
ICU dialysis is not limited to adults. Newborns and infants occasionally develop kidney failure, particularly in the setting of congenital heart disease, sepsis, or complications of cardiac surgery. The challenge is that standard dialysis equipment was designed for adult-sized blood volumes and flow rates. Running a newborn’s blood through an adult-sized circuit can cause dangerous drops in blood pressure and body temperature. Specialized miniaturized systems have been developed for this population. One such device, designed with a circuit volume of less than 30 milliliters and highly precise fluid-removal control, was used to successfully treat a 2.9-kilogram neonate with hemorrhagic shock, multi-organ failure, and severe fluid overload using multiple modalities over more than 400 hours. The infant survived and was discharged with only mild residual kidney impairment.28The Lancet. Extracorporeal membrane oxygenation or renal replacement therapy in neonates These cases remain rare and technically demanding, but they illustrate how far the technology has come in extending life-saving kidney support across the age spectrum.