Acute Kidney Injury & CRRT

Acute Kidney Injury & CRRT: 2026 Evidence & Management

Current standard of care for acute kidney injury in the critically ill — diagnosis, prevention, fluids, and kidney replacement therapy — organized by treatment area, with the trials behind each recommendation.

Reviewed by the DosePilot Medical Team · Last reviewed: Jun 2026 · Independent — no industry funding
How to read this page. Each area opens with the current approach, then the trials behind it (design, key results, limitations), and a short practical note (In practice). All thresholds, doses, and figures are taken from the cited primary sources (DOI + PMID). The KDIGO 2012 framework remains the reference standard; a KDIGO AKI/AKD update is in public review, and 2025 ICU guidance (SRLF–GFRUP) and a current Lancet seminar are incorporated where relevant. In practice notes are general guidance, not a substitute for clinical judgment.
How we verify evidence & editorial policy
Written and maintained by the DosePilot Medical Team. Recommendations are based on randomized controlled trials and current society guidelines (GRADE). Every citation is verified directly against PubMed (PMID) and the publisher DOI — no memory-based or AI-generated references. Each topic is peer-reviewed by the clinical team before publication and re-reviewed at least every 12 months or when major new evidence appears.

Funding & conflicts of interest. Produced independently; no industry funding was received for this content and contributors report no relevant financial conflicts of interest.
On this page
  1. Evidence timeline
  2. Definition, staging & diagnosis
  3. Prevention & nephroprotection
  4. Fluids & hemodynamics
  5. Diuretics & medical management
  6. When to start kidney replacement therapy
  7. Modality — CRRT, IHD or SLED
  8. CRRT prescription — dose & anticoagulation
  9. Fluid removal & net ultrafiltration
  10. Drug dosing during kidney replacement
  11. Weaning from RRT & recovery
  12. Where it’s heading
  13. Bottom line
  14. FAQ
  15. Related DosePilot tools
  16. Guidelines & evidence basis
At a glance — current standard of care
AreaCurrent approachKey source(s)
DefinitionSCr ↑ ≥0.3 mg/dL in 48 h, or ≥1.5× baseline in 7 d, or UO <0.5 mL/kg/h for ≥6 hKDIGO 2012
BiomarkersTIMP-2·IGFBP7 / NGAL flag risk before creatinine risesSapphire
PreventionKDIGO care bundle in high-risk; avoid nephrotoxins; perfusionPrevAKI; BigpAK-2
ContrastVolume expansion; NAC & bicarbonate no benefitPRESERVE
FluidsBalanced crystalloids ≥ saline; avoid fluid overloadSMART; PLUS
AreaCurrent approachKey trial(s)
When to startNo routine early start; treat refractory complications; don’t over-delaySTARRT-AKI; AKIKI-2
ModalityCRRT, IHD, SLED equivalent for survival; CRRT/SLED if unstableHemodiafe; meta
DoseEffluent 20–25 mL/kg/h delivered (prescribe higher)RENAL; ATN
AnticoagulationRegional citrate first-line (longer filter life)RICH
Fluid removalModerate net ultrafiltration; avoid high-intensity removalRENAL 2° analysis

1Evidence timeline (2004–2026)

2004RIFLE (ADQI) — the first consensus definition and staging of acute kidney injury
2007AKIN — refines the criteria, adding the small absolute creatinine rise (≥0.3 mg/dL in 48 h)
2008ATN — higher-intensity renal support does not improve survival
2009RENAL — confirms standard-dose CRRT (25 mL/kg/h) equals higher dose; sets the dosing standard
2012KDIGO — unifies the AKI definition and staging and introduces the supportive “care bundle”
2013Cell-cycle-arrest biomarkers (TIMP-2·IGFBP7, Sapphire) and the furosemide stress test add early risk stratification
2016ELAIN vs AKIKI — conflicting results on early versus delayed start of kidney replacement therapy
2017PrevAKI — a biomarker-guided KDIGO care bundle reduces AKI after cardiac surgery
2018IDEAL-ICU — no benefit of early RRT in septic shock; PRESERVE — bicarbonate & N-acetylcysteine fail for contrast AKI; SMART — balanced crystalloids reduce adverse kidney events
2020STARRT-AKI — accelerated RRT gives no survival benefit and more dialysis dependence; RICH — regional citrate prolongs filter life; Gaudry IPD meta — delayed ≈ early
2021AKIKI-2 — waiting too long to start RRT is associated with harm; BaSICS — balanced ≈ saline
2022PLUS — balanced multielectrolyte solution ≈ saline in a large general-ICU trial
2025SRLF–GFRUP ICU renal-replacement guideline and a current Lancet AKI seminar; BigpAK-2 — a biomarker-guided care bundle cuts moderate-to-severe post-operative AKI
2025–26A KDIGO AKI/AKD update is in progress; biomarker-guided prevention and precision fluid removal move toward the bedside — no new mortality-changing modality or dose

2Definition, staging & diagnosis

Current approach (2026)
Consensus: KDIGO stagingBiomarkers refine risk
Diagnose AKI by the KDIGO criteria: a rise in serum creatinine of ≥0.3 mg/dL within 48 h, a rise to ≥1.5× baseline within 7 days, or urine output <0.5 mL/kg/h for ≥6 h. Stage 1–3 by the worst of the creatinine or urine-output criterion. Establish a baseline creatinine, then work through the differential — pre-renal (hypoperfusion), intrinsic (acute tubular injury, most often sepsis- or ischemia-associated; also nephrotoxic and glomerular/interstitial), and post-renal (obstruction) — with a focused history, medication review, urinalysis with microscopy, and a bladder/renal ultrasound to exclude obstruction. (If you use the fractional excretion of sodium to separate pre-renal from intrinsic AKI, remember it is uninterpretable on diuretics — use the fractional excretion of urea instead.) Damage biomarkers (urinary TIMP-2·IGFBP7, NGAL) and the furosemide stress test can flag risk or predict progression before creatinine moves. Think in terms of the AKI → acute kidney disease (AKD) → CKD continuum.
KDIGO stageSerum creatinineUrine output
11.5–1.9× baseline, or ≥0.3 mg/dL rise<0.5 mL/kg/h for 6–12 h
22.0–2.9× baseline<0.5 mL/kg/h for ≥12 h
3≥3.0× baseline, or ≥4.0 mg/dL, or started on RRT<0.3 mL/kg/h for ≥24 h, or anuria ≥12 h
Evidence — old → new
2004
RIFLE (ADQI). The first consensus framework graded kidney injury by creatinine/GFR and urine output (Risk–Injury–Failure–Loss–ESKD), showing that even small changes track with mortality and standardizing research definitions.
2007
AKIN. Refined RIFLE into a 48-hour staging system and added the small absolute rise (≥0.3 mg/dL), recognizing that minor creatinine increases carry prognostic weight.
2012
KDIGO. Merged RIFLE and AKIN into a single, now-universal definition and 3-stage system, and framed supportive management (the “KDIGO bundle”). A KDIGO AKI/AKD update is in public review but the 2012 criteria remain the reference standard.
2013
Cell-cycle-arrest biomarkers (Sapphire). Urinary [TIMP-2]·[IGFBP7] identified patients who would develop moderate-to-severe AKI within 12 h better than older markers (AUC ≈ 0.80). The furosemide stress test (1.0–1.5 mg/kg, then measure 2-h urine output) predicts progression to stage 3 and need for RRT.

What changed. AKI moved from a dozen competing definitions to one consensus system, and diagnosis is shifting from a functional marker (creatinine, which lags injury by 24–48 h) toward earlier damage/stress biomarkers and dynamic tests — while still requiring clinical context.

In practice

Stage by the worse of creatinine or urine output, and always anchor to a true baseline. Early: confirm volume/perfusion, scan for obstruction, and stop nephrotoxins. Creatinine trails the injury — a normal value early does not exclude evolving AKI; a biomarker or furosemide stress test can help when the trajectory is unclear.

Sources (5)
  • Bellomo R, Ronco C, Kellum JA, et al; ADQI workgroup. Acute renal failure — definition, outcome measures, animal models, fluid therapy and information technology needs: the RIFLE criteria. Crit Care. 2004;8(4):R204–R212. DOI 10.1186/cc2872. PMID 15312219.
  • Mehta RL, Kellum JA, Shah SV, et al; Acute Kidney Injury Network. AKIN: report of an initiative to improve outcomes in acute kidney injury. Crit Care. 2007;11(2):R31. DOI 10.1186/cc5713. PMID 17331245.
  • Kellum JA, Lameire N; KDIGO AKI Guideline Work Group. Diagnosis, evaluation, and management of acute kidney injury: a KDIGO summary (Part 1). Crit Care. 2013;17(1):204. DOI 10.1186/cc11454. PMID 23394211.
  • Kashani K, Al-Khafaji A, Ardiles T, et al. Discovery and validation of cell cycle arrest biomarkers in human acute kidney injury (Sapphire). Crit Care. 2013;17(1):R25. DOI 10.1186/cc12503. PMID 23388612.
  • Chawla LS, Davison DL, Brasha-Mitchell E, et al. Development and standardization of a furosemide stress test to predict the severity of acute kidney injury. Crit Care. 2013;17(5):R207. DOI 10.1186/cc13015. PMID 24053972.

3Prevention & nephroprotection

Current approach (2026)
RCT: biomarker-guided KDIGO bundle reduces AKI (PrevAKI, BigpAK-2)RCT: NAC & bicarbonate no benefit (PRESERVE)
There is no nephroprotective drug; prevention is a process. In high-risk patients — major or cardiac surgery, sepsis, or a positive stress biomarker — implement the KDIGO care bundle: optimize volume status and hemodynamics, maintain perfusion pressure, use functional hemodynamic monitoring, discontinue nephrotoxic agents where possible, avoid hyperglycemia, and minimize/replace iodinated contrast. For contrast exposure in higher-risk patients (e.g., eGFR <30, or 30–45 with added risk factors), volume expansion with crystalloid is the only proven measure — N-acetylcysteine and sodium bicarbonate do not help, and most patients with preserved kidney function need no specific prophylaxis. “Renal-dose” dopamine, fenoldopam, and natriuretic peptides are not recommended for prevention.
Evidence — old → new
2017
PrevAKI. Single-center RCT, 276 cardiac-surgery patients with urinary [TIMP-2]·[IGFBP7] >0.3; full KDIGO bundle vs usual care. The bundle reduced AKI within 72 h (55.1% vs 71.7%) and lowered moderate-to-severe AKI, with far better adherence to the full bundle in the intervention arm.
2018
PRESERVE. RCT, 5,177 high-risk patients before angiography, 2×2 design: IV sodium bicarbonate vs saline and oral N-acetylcysteine vs placebo. No benefit on the composite of death, dialysis, or persistent kidney decline at 90 days (4.4% vs 4.7%) or on contrast-associated AKI. Hydration with either fluid is the active ingredient.
2025
BigpAK-2. Multinational RCT, 34 European hospitals; biomarker-identified high-risk patients after major surgery randomized to the KDIGO care bundle vs standard care. Among 1,176 patients analyzed, moderate-to-severe AKI within 72 h fell to 14.4% vs 22.3% (OR 0.57, 95% CI 0.40–0.79; NNT 12) with no excess adverse events — confirming PrevAKI in a large, general surgical population.

What changed. A generation of “magic-bullet” prophylactic drugs (dopamine, fenoldopam, NAC, statins, natriuretic peptides) failed; what works is a disciplined supportive bundle delivered to the right (biomarker-identified) patients. Contrast nephropathy has also proven less common and less preventable-by-drug than once thought.

In practice

Flag high-risk patients (cardiac/major surgery, sepsis, biomarker-positive) and hard-wire the bundle: euvolemia and perfusion pressure, daily nephrotoxin review (NSAIDs, aminoglycosides, contrast load), and glucose control. Before contrast: hydrate with crystalloid, use the lowest necessary volume, and skip NAC/bicarbonate.

Sources (3)
  • Meersch M, Schmidt C, Hoffmeier A, et al. Prevention of cardiac surgery-associated AKI by implementing the KDIGO guidelines in high-risk patients identified by biomarkers: the PrevAKI randomized controlled trial. Intensive Care Med. 2017;43(11):1551–1561. DOI 10.1007/s00134-016-4670-3. PMID 28110412.
  • Weisbord SD, Gallagher M, Jneid H, et al; PRESERVE Trial Group. Outcomes after angiography with sodium bicarbonate and acetylcysteine. N Engl J Med. 2018;378(7):603–614. DOI 10.1056/NEJMoa1710933. PMID 29130810.
  • Zarbock A, Küllmar M, Ostermann M, et al. A preventive care strategy to reduce moderate or severe acute kidney injury after major surgery (BigpAK-2): a multinational, randomised clinical trial. Lancet. 2025;406(10521):2782–2791. DOI 10.1016/S0140-6736(25)01717-9. PMID 41242333.

4Fluids & hemodynamics

Current approach (2026)
RCT: balanced ≥ saline for kidney events (SMART)RCT: balanced ≈ saline overall (BaSICS, PLUS)
Resuscitate to perfusion, not to a fixed number — use dynamic measures of fluid responsiveness and reassess often. Prefer balanced crystalloids (Ringer’s lactate, Plasma-Lyte) as the default; large volumes of 0.9% saline cause hyperchloremic acidosis and may modestly worsen kidney outcomes. Hold mean arterial pressure with vasopressors rather than chasing ever-more fluid (target MAP ≥65 mmHg; a higher target may reduce RRT in chronic hypertension — SEPSISPAM). Critically, avoid fluid overload: a positive cumulative balance is independently associated with AKI and death, so de-resuscitate (stop maintenance fluids, remove excess) once the patient is stabilized.
Evidence — old → new
2018
SMART. Pragmatic trial, 15,802 ICU adults, balanced crystalloids vs saline. Major adverse kidney events at 30 days (death, new RRT, or persistent renal dysfunction) were 14.3% vs 15.4% (OR 0.91, 95% CI 0.84–0.99) — a small but significant benefit, most evident in sepsis and larger-volume use.
2021
BaSICS. RCT, 11,052 ICU patients in Brazil, balanced vs saline. No difference in 90-day mortality (26.4% vs 27.2%) or AKI/RRT overall, though most patients received modest fluid volumes.
2022
PLUS. RCT, 5,037 critically ill adults in Australia/New Zealand, balanced multielectrolyte solution vs saline. No difference in 90-day mortality (21.8% vs 22.0%) or new RRT. Meta-analyses including SMART, BaSICS and PLUS suggest a small probability of benefit favoring balanced fluids.

What changed. Two shifts: from saline-by-default to balanced-by-default (cheap, safe, plausibly better for the kidney), and from liberal resuscitation to limiting and then actively removing excess fluid, recognizing overload itself as an injury.

In practice

Use balanced crystalloid for resuscitation and maintenance; reserve saline for specific indications (hypochloremia, hyponatremia, certain neuro cases). Give fluid only to a responsive patient, then switch to vasopressors to hold MAP. Track cumulative balance and de-resuscitate early — daily “can I take fluid off?” matters as much as how much you put in.

Sources (4)
  • Semler MW, Self WH, Wanderer JP, et al; SMART Investigators. Balanced crystalloids versus saline in critically ill adults. N Engl J Med. 2018;378(9):829–839. DOI 10.1056/NEJMoa1711584. PMID 29485925.
  • Zampieri FG, Machado FR, Biondi RS, et al; BaSICS Investigators. Effect of intravenous fluid treatment with a balanced solution vs 0.9% saline solution on mortality in critically ill patients: the BaSICS randomized clinical trial. JAMA. 2021;326(9):818–829. DOI 10.1001/jama.2021.11684. PMID 34375394.
  • Finfer S, Micallef S, Hammond N, et al; PLUS Study Investigators. Balanced multielectrolyte solution versus saline in critically ill adults. N Engl J Med. 2022;386(9):815–826. DOI 10.1056/NEJMoa2114464. PMID 35041780.
  • Asfar P, Meziani F, Hamel JF, et al; SEPSISPAM Investigators. High versus low blood-pressure target in patients with septic shock. N Engl J Med. 2014;370(17):1583–1593. DOI 10.1056/NEJMoa1312173. PMID 24635770.

5Diuretics & medical management

Current approach (2026)
Meta: loop diuretics don’t prevent/treat AKIRCT subgroup: bicarbonate in severe acidemia + AKI (BICAR-ICU)
Loop diuretics do not prevent or treat AKI and do not improve survival or kidney recovery — use them only to manage fluid overload in a diuretic-responsive patient, never to “convert” oliguric to non-oliguric AKI or to postpone clearly indicated RRT. Manage the medical complications that otherwise drive RRT: treat hyperkalemia medically (calcium, insulin–glucose, β-agonist, and potassium removal), correct severe metabolic acidosis — IV sodium bicarbonate may reduce the need for RRT in severe acidemia with AKI — and relentlessly remove nephrotoxins. Dopamine, fenoldopam, and mannitol have no role in AKI.
Evidence — old → new
2006
Furosemide meta-analysis (Ho & Sheridan). Pooled RCTs found furosemide did not reduce mortality, need for RRT, or the number of dialysis sessions in AKI, and high doses risked ototoxicity (limit IV infusion to ≈4 mg/min; additive risk with aminoglycosides). Diuretics treat the symptom (volume), not the disease.
2018
BICAR-ICU. RCT, 389 ICU patients with severe metabolic acidemia (pH ≤7.20). Bicarbonate showed no overall benefit, but in the prespecified moderate-to-severe AKI (AKIN 2–3) stratum it reduced the primary composite and the need for RRT — a hypothesis-generating subgroup signal, not a definitive result.

What changed. Loop diuretics were demoted from “renal protection” to volume management only, and bicarbonate found a narrow, evidence-based niche in severe acidemia with AKI. The list of disproven nephroprotective drugs keeps growing.

In practice

Use furosemide to offload a fluid-overloaded, responsive patient (and gauge responsiveness) — not as a kidney “cure” and not to stall a needed circuit. Treat hyperkalemia and acidosis on their merits; in pH ≤7.20 with moderate-to-severe AKI, bicarbonate may reduce the need for dialysis. Re-list and stop nephrotoxins daily.

Sources (2)
  • Ho KM, Sheridan DJ. Meta-analysis of frusemide to prevent or treat acute renal failure. BMJ. 2006;333(7565):420. DOI 10.1136/bmj.38902.605347.7C. PMID 16861256.
  • Jaber S, Paugam C, Futier E, et al; BICAR-ICU Study Group. Sodium bicarbonate therapy for patients with severe metabolic acidaemia in the intensive care unit (BICAR-ICU): a multicentre, open-label, randomised controlled, phase 3 trial. Lancet. 2018;392(10141):31–40. DOI 10.1016/S0140-6736(18)31080-8. PMID 29910040.

6When to start kidney replacement therapy

Current approach (2026)
RCT: accelerated start — no benefit (STARRT-AKI)RCT: over-delay — harm (AKIKI-2)
Start RRT urgently for a refractory, life-threatening complication: refractory hyperkalemia, severe metabolic acidosis, refractory fluid overload / pulmonary edema, or overt uremic complications (encephalopathy, pericarditis) — plus certain dialyzable intoxications. In the absence of these emergencies, there is no benefit to a routine “early” or “accelerated” start: watchful waiting spares many patients dialysis entirely and avoids catheter and circuit harms. But do not over-delay — once severe AKI is prolonged (e.g., oliguria >72 h or markedly rising urea) or a clear indication appears, waiting longer is associated with harm. Decide by trajectory and complications, not by stage or a creatinine number alone.
Evidence — old → new
2016
ELAIN. Single-center RCT, 231 mostly surgical patients; early (KDIGO 2) vs delayed (KDIGO 3). Early start lowered 90-day mortality (39.3% vs 54.7%; HR 0.66) — the lone positive timing trial, limited by single-center, surgical population.
2016
AKIKI. Multicenter RCT, 620 patients; early (KDIGO 3) vs delayed (await urgent indication). No difference in 60-day mortality (48.5% vs 49.7%), and 49% of the delayed group never needed RRT; more catheter-related bloodstream infections with early start.
2018
IDEAL-ICU. RCT, 488 patients with septic shock and severe AKI; early (within 12 h) vs delayed (48 h). No mortality difference (90-day 58% vs 54%); stopped for futility — early start gave no advantage even in septic AKI.
2020
STARRT-AKI. The definitive trial: 3,019 patients (modified ITT 2,927), accelerated vs standard initiation. 90-day mortality identical (43.9% vs 43.7%), but accelerated start caused more RRT dependence at 90 days (10.4% vs 6.0%) and more adverse events. No survival benefit to starting early.
2020
Gaudry individual-patient-data meta-analysis. Pooling the major timing trials, a delayed strategy gave the same 28-day survival as early initiation while reducing RRT exposure — consolidating “watchful waiting” as the default.
2021
AKIKI-2. RCT, 278 patients already meeting delayed criteria; delayed vs more-delayed. A longer wait did not add RRT-free days (median 12 vs 10; p=0.93) and carried a higher 60-day mortality (adjusted HR 1.65, 95% CI 1.09–2.50) — defining the lower limit of safe delay.

What changed. The pendulum swung from “earlier is better” (ELAIN) to a consistent message from larger trials — no benefit, and some harm, from pre-emptive RRT — and then AKIKI-2 set the floor: watchful waiting is right, but indefinite delay is dangerous. Start for an indication, not a stage.

In practice

Ask “is there a refractory complication now?” — if yes, start. If no, watch closely (potassium, pH, volume, urea, urine output) and start when an indication emerges or severe AKI persists. Don’t accelerate a stable patient, and don’t let a deteriorating one drift past the point AKIKI-2 flagged as harmful.

Sources (6)
  • Zarbock A, Kellum JA, Schmidt C, et al. Effect of early vs delayed initiation of renal replacement therapy on mortality in critically ill patients with acute kidney injury: the ELAIN randomized clinical trial. JAMA. 2016;315(20):2190–2199. DOI 10.1001/jama.2016.5828. PMID 27209269.
  • Gaudry S, Hajage D, Schortgen F, et al; AKIKI Study Group. Initiation strategies for renal-replacement therapy in the intensive care unit. N Engl J Med. 2016;375(2):122–133. DOI 10.1056/NEJMoa1603017. PMID 27181456.
  • Barbar SD, Clere-Jehl R, Bourredjem A, et al; IDEAL-ICU Trial Investigators. Timing of renal-replacement therapy in patients with acute kidney injury and sepsis. N Engl J Med. 2018;379(15):1431–1442. DOI 10.1056/NEJMoa1803213. PMID 30304656.
  • STARRT-AKI Investigators. Timing of initiation of renal-replacement therapy in acute kidney injury. N Engl J Med. 2020;383(3):240–251. DOI 10.1056/NEJMoa2000741. PMID 32668114.
  • Gaudry S, Hajage D, Benichou N, et al. Delayed versus early initiation of renal replacement therapy for severe acute kidney injury: a systematic review and individual patient data meta-analysis of randomised clinical trials. Lancet. 2020;395(10235):1506–1515. DOI 10.1016/S0140-6736(20)30531-6. PMID 32334654.
  • Gaudry S, Hajage D, Martin-Lefevre L, et al. Comparison of two delayed strategies for renal replacement therapy initiation for severe acute kidney injury (AKIKI 2): a multicentre, open-label, randomised, controlled trial. Lancet. 2021;397(10281):1293–1300. DOI 10.1016/S0140-6736(21)00350-0. PMID 33812488.

7Modality — CRRT, IHD or SLED

Current approach (2026)
RCT/meta: no modality superior for survival
No modality improves survival or kidney recovery over another — choose by hemodynamics and the clinical goal. Use continuous RRT (CRRT) — or prolonged intermittent / sustained low-efficiency dialysis (SLED/PIRRT) — in hemodynamically unstable patients, where gradual solute and fluid removal is better tolerated. Use intermittent hemodialysis (IHD) in stable patients and when rapid clearance is needed (severe hyperkalemia, dialyzable poisons) or to free a recovering patient for mobilization. SLED/PIRRT is a practical hybrid. Transition between modalities as the patient’s stability changes.
Evidence — old → new
2006
Hemodiafe. Multicenter RCT, 360 patients with AKI and multi-organ failure; CVVHDF vs IHD (with protocols allowing IHD even in shock). No difference in 60-day survival, showing IHD can be delivered safely with attention to tolerance.
2008
CRRT vs IHD meta-analyses. Pooled RCTs find no difference in mortality or length of stay. CRRT provides greater hemodynamic stability and, in observational data (inconsistent across studies), may favor kidney recovery / less long-term dialysis dependence; SLED cohorts show similar mortality with a slightly higher risk of persistent RRT dependence. All three are reasonable when matched to the patient.

What changed. Early enthusiasm that CRRT would cut mortality gave way to repeated demonstrations of equivalence; the decision now turns on hemodynamic tolerance, clearance speed, mobilization, and local resources rather than a survival argument.

In practice

On pressors or hemodynamically marginal → CRRT (or SLED). Stable, or needing fast potassium/toxin clearance → IHD. Use SLED to balance tolerance with nursing/throughput. Don’t keep a stabilizing patient on continuous therapy out of habit — step down to free them up.

Sources (2)
  • Vinsonneau C, Camus C, Combes A, et al; Hemodiafe Study Group. Continuous venovenous haemodiafiltration versus intermittent haemodialysis for acute renal failure in patients with multiple-organ dysfunction syndrome: a multicentre randomised trial. Lancet. 2006;368(9533):379–385. DOI 10.1016/S0140-6736(06)69111-3. PMID 16876666.
  • Bagshaw SM, Berthiaume LR, Delaney A, Bellomo R. Continuous versus intermittent renal replacement therapy for critically ill patients with acute kidney injury: a meta-analysis. Crit Care Med. 2008;36(2):610–617. DOI 10.1097/01.CCM.0B013E3181611F552. PMID 18216610.

8CRRT prescription — dose & anticoagulation

Current approach (2026)
RCT: higher dose — no benefit (RENAL, ATN)RCT: regional citrate first-line (RICH)
Prescribe a delivered effluent dose of 20–25 mL/kg/h; because circuit downtime erodes delivery, prescribe ~25–30 mL/kg/h to actually deliver ~25. Higher doses do not improve outcomes and increase hypophosphatemia and antibiotic/nutrient losses. Convective and diffusive clearance (CVVH, CVVHD, CVVHDF) are equivalent — choose by local practice. For anticoagulation, regional citrate is first-line over systemic heparin (longer filter life, fewer bleeding events). Citrate chelates ionized calcium in the circuit, so protocols require frequent systemic ionized-calcium monitoring with calcium replacement; watch for metabolic alkalosis (citrate is metabolized to bicarbonate) and, in severe liver failure or deep shock, for citrate accumulation (metabolic acidosis with a rising total-to-ionized calcium ratio >2.5 and rising calcium requirements). Use heparin or anticoagulant-free circuits when citrate is unavailable or contraindicated, and secure good vascular access (right internal jugular preferred, then femoral; avoid the subclavian vein to preserve future fistula sites).
Evidence — old → new
2008
ATN. RCT, 1,124 patients; intensive vs less-intensive support (CRRT 35 vs 20 mL/kg/h; IHD/SLED 6× vs 3×/week). No difference in 60-day mortality (53.6% vs 51.5%) or recovery — more is not better.
2009
RENAL. RCT, 1,508 patients; CVVHDF at 40 vs 25 mL/kg/h. Identical 90-day mortality (44.7% in both), with more hypophosphatemia in the high-dose arm. Together with ATN this fixed the dose target at ~20–25 mL/kg/h delivered.
2020
RICH. RCT, 638 patients randomized on CRRT; regional citrate vs systemic heparin. Citrate markedly prolonged filter life span (median 47 vs 26 h) and reduced bleeding; 90-day mortality was not significantly different (trial stopped early). Metabolic effects of citrate were manageable with protocols.

What changed. Dose intensity was settled by two large RCTs (no benefit beyond ~25 mL/kg/h delivered), and anticoagulation shifted from heparin to regional citrate as the default, now endorsed by guidelines for most patients without a citrate contraindication.

In practice

Write the order for ~25–30 mL/kg/h to deliver ~25 after downtime; track the actual delivered dose. Default to regional citrate; in liver failure or refractory shock, watch the total:ionized calcium ratio and ionized calcium for accumulation. Replace phosphate, and re-check antibiotic dosing (§10). Minimize circuit interruptions and use a well-positioned catheter.

Sources (3)
  • VA/NIH Acute Renal Failure Trial Network; Palevsky PM, Zhang JH, O’Connor TZ, et al. Intensity of renal support in critically ill patients with acute kidney injury (ATN). N Engl J Med. 2008;359(1):7–20. DOI 10.1056/NEJMoa0802639. PMID 18492867.
  • RENAL Replacement Therapy Study Investigators; Bellomo R, Cass A, Cole L, et al. Intensity of continuous renal-replacement therapy in critically ill patients. N Engl J Med. 2009;361(17):1627–1638. DOI 10.1056/NEJMoa0902413. PMID 19846848.
  • Zarbock A, Küllmar M, Kindgen-Milles D, et al; RICH Investigators. 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 AKI: the RICH randomized clinical trial. JAMA. 2020;324(16):1629–1639. DOI 10.1001/jama.2020.18618. PMID 33095849.

9Fluid removal & net ultrafiltration

Current approach (2026)
Observational: moderate net ultrafiltration; avoid extremes
Once resuscitation is complete, use kidney replacement to correct fluid overload deliberately — but at a moderate net ultrafiltration (NUF) rate. Both too little removal (persistent overload, itself linked to death) and too aggressive removal (hemodynamic instability, organ hypoperfusion) are harmful. Observational data suggest a window around ~1.0–1.5 mL/kg/h, with high-intensity removal (>1.75 mL/kg/h) associated with higher mortality. Set a daily fluid-balance goal, titrate NUF to hemodynamics, and remember that many instability episodes on CRRT are not preload-related — check responsiveness before reflexively stopping fluid removal.
Evidence
2019
Net ultrafiltration & mortality (RENAL secondary analysis, Murugan). Among CRRT patients, high-intensity NUF (>1.75 mL/kg/h) was associated with higher 90-day mortality compared with moderate rates (1.01–1.75), while very low rates risked sustained overload — describing a therapeutic window rather than “more off is better.”

What changed. Fluid removal is now treated as a titratable exposure with its own dose–response, not an afterthought. “Precision net ultrafiltration” — individualizing the rate to hemodynamics — is an active area, with randomized trials (e.g., fluid-balance–neutral strategies) underway.

In practice

After resuscitation, remove fluid actively but moderately (≈1–1.5 mL/kg/h), set a daily balance target, and reassess tolerance often. Don’t halt ultrafiltration for every blood-pressure dip — confirm whether the patient is actually preload-responsive first. Persistent overload is as dangerous as over-removal.

Sources (1)
  • Murugan R, Kerti SJ, Chang CH, et al. Association of net ultrafiltration rate with mortality among critically ill adults with acute kidney injury receiving continuous venovenous hemodiafiltration: a secondary analysis of the RENAL trial. JAMA Netw Open. 2019;2(6):e195418. DOI 10.1001/jamanetworkopen.2019.5418. PMID 31173127.

10Drug dosing during kidney replacement

Current approach (2026)
Guidance: full loading dose; TDMUnderdosing worsens outcomes
Antibiotic underdosing is common and dangerous on kidney replacement — give the full loading dose (it depends on volume of distribution, not clearance — do not reduce it), then set maintenance by the modality, delivered effluent/clearance, and residual renal function. Hydrophilic agents (β-lactams, vancomycin, aminoglycosides) are substantially removed by CRRT: increase maintenance dosing for the time-/AUC-dependent agents (β-lactams, vancomycin), while for concentration-dependent aminoglycosides give a full peak dose and then extend the interval, redosing by trough-guided TDM; lipophilic drugs are cleared less. Use therapeutic drug monitoring (TDM) wherever available, re-dose dialyzable drugs after IHD sessions, and remember that ICU patients without AKI can have augmented renal clearance and need higher doses. Involve a critical-care pharmacist.
Evidence
—
Underdosing is the rule, not the exception. Point-prevalence and cohort studies show a large fraction of CRRT patients are underexposed to β-lactams early in therapy, and antibiotic underexposure is linked to worse clinical outcomes in sepsis. TDM frequently identifies a need to change the β-lactam dose in CRRT patients — more often up than down.

What changed. The old reflex — “kidney failure, so cut every dose” — proved harmful for anti-infectives during CRRT, which clears these drugs efficiently. The modern rule is full loading dose, clearance-matched maintenance, and TDM, prioritizing early adequate exposure in sepsis.

In practice

Never trim the loading dose. Set β-lactam/vancomycin maintenance to the CRRT modality and delivered dose (and any residual urine output), use TDM when you can, and time doses around IHD. Re-check whenever the RRT prescription or kidney function changes. A pharmacist-driven CRRT dosing protocol prevents both under- and over-exposure.

Sources (2)
  • Joint guideline. Jourdain M, Chatti K, Klouche K, et al. Renal replacement therapy in an intensive care unit: guidelines from the SRLF–GFRUP consensus conference. Ann Intensive Care. 2025;15(1):100. DOI 10.1186/s13613-025-01517-0. PMID 40668437.
  • Pistolesi V, Morabito S, Di Mario F, et al. A guide to understanding antimicrobial drug dosing in critically ill patients on renal replacement therapy. Antimicrob Agents Chemother. 2019;63(8):e00583-19. DOI 10.1128/AAC.00583-19. PMID 31109983.

11Weaning from RRT & recovery

Current approach (2026)
Observational: urine output predicts successful cessation
Reassess daily for kidney recovery and stop RRT once the kidney can keep up — the strongest practical signal is a rising spontaneous urine output (off diuretics), supported by an improving measured clearance; there is no single validated threshold, so judge by trajectory and reduced need for solute/fluid control. After AKI, recognize the acute kidney disease (AKD) phase: avoid nephrotoxins, recheck kidney function, reconcile chronic medications (including careful re-introduction of RAAS inhibitors), and arrange nephrology follow-up for survivors — AKI substantially raises the long-term risk of CKD, recurrent AKI, cardiovascular events, and death.
Evidence — old → new
2009
Predictors of successful cessation (BEST Kidney). In a large observational CRRT cohort, urine output at the time of stopping was the best predictor of remaining RRT-free (more predictive than creatinine or urea), informing today’s “watch the urine” approach to weaning.
2025
The AKI → CKD continuum (Lancet seminar). Contemporary syntheses emphasize that AKI survivors carry markedly higher risks of CKD, recurrent AKI and mortality, and that structured follow-up — function recheck, nephrotoxin avoidance, medication review, and risk-based nephrology referral — closes a major care gap.

What changed. Discontinuation became evidence-informed (urine output as the key cue), and the field reframed AKI as the start of a longitudinal kidney problem rather than a self-limited ICU event — making post-AKI follow-up a quality metric.

In practice

Each day, ask “can RRT stop?” — a recovering, off-diuretic urine output is the green light; trial cessation rather than continuing by inertia. At discharge, document AKD, list it for follow-up creatinine, avoid nephrotoxins, restart chronic meds thoughtfully, and refer higher-risk survivors to nephrology.

Sources (2)
  • Uchino S, Bellomo R, Morimatsu H, et al. Discontinuation of continuous renal replacement therapy: a post hoc analysis of a prospective multicenter observational study (BEST Kidney). Crit Care Med. 2009;37(9):2576–2582. DOI 10.1097/CCM.0b013e3181a38241. PMID 19623048.
  • Ostermann M, Lumlertgul N, Jeong R, See E, Joannidis M, James M. Acute kidney injury. Lancet. 2025;405(10474):241–256. DOI 10.1016/S0140-6736(24)02385-7. PMID 39826969.

12Where it’s heading (2026)

A KDIGO AKI/AKD update. A revised guideline is in public review (final pending), expected to formalize the AKI–AKD–CKD continuum and incorporate damage/stress biomarkers into definition and risk stratification.

Sub-phenotyping & precision care. Distinguishing functional from structural injury, and identifying treatment-responsive subgroups (for example, the angiotensin-II–responsive vasodilatory-shock-with-AKI signal from ATHOS-3), aims to move beyond one-size-fits-all support.

Precision fluid removal & novel circuits. Randomized trials of fluid-balance–neutral / individualized net ultrafiltration, the selective cytopheretic device (SCD) as an adjunct to CRRT, and non-heparin anticoagulants (e.g., nafamostat) are in progress.

Earlier detection. Machine-learning prediction and electronic AKI alerts, coupled with biomarker-guided bundles (PrevAKI, BigpAK-2), are shifting effort upstream toward prevention.

13Bottom line — the AKI workflow

  1. Diagnose & stage by KDIGO (worse of creatinine or urine output); anchor to a baseline, exclude obstruction and pre-renal causes, and stop nephrotoxins.
  2. Prevent in high-risk patients with the KDIGO care bundle (PrevAKI, BigpAK-2); for contrast, hydrate and skip NAC/bicarbonate (PRESERVE).
  3. Resuscitate with balanced crystalloids to perfusion (SMART), then de-resuscitate — avoid fluid overload; hold MAP with vasopressors.
  4. Diuretics for volume only; bicarbonate for severe acidemia with moderate-to-severe AKI (BICAR-ICU subgroup); no dopamine/fenoldopam/mannitol.
  5. Start RRT for a refractory complication (hyperkalemia, acidosis, overload, uremia); otherwise watchful waiting (STARRT-AKI) — but don’t over-delay (AKIKI-2).
  6. Modality by hemodynamics — CRRT/SLED if unstable, IHD if stable or for fast clearance; all equal for survival.
  7. CRRT dose ~25 mL/kg/h delivered (RENAL, ATN); regional citrate first-line (RICH).
  8. Net ultrafiltration moderate (~1–1.5 mL/kg/h) — avoid both overload and over-removal.
  9. Don’t underdose antibiotics — full loading dose plus clearance-matched maintenance and TDM.
  10. Reassess recovery daily (watch urine output); after AKI, label AKD, avoid nephrotoxins, and arrange nephrology follow-up.

14FAQ

How is acute kidney injury defined?
By the KDIGO criteria: a rise in serum creatinine of ≥0.3 mg/dL within 48 hours, a rise to ≥1.5 times baseline within 7 days, or urine output below 0.5 mL/kg/h for at least 6 hours. Stage 1–3 is assigned by whichever criterion (creatinine or urine output) is worse.
When should dialysis be started in AKI?
Urgently for a refractory, life-threatening complication — severe hyperkalemia, severe acidosis, refractory fluid overload, or uremic complications. Otherwise there is no benefit to a routine early start (STARRT-AKI); watchful waiting spares many patients dialysis. But do not delay indefinitely, because waiting too long was associated with harm (AKIKI-2).
Is CRRT better than intermittent hemodialysis?
No modality improves survival or kidney recovery over another. Continuous RRT (or SLED) is preferred when the patient is hemodynamically unstable because fluid and solutes are removed gradually; intermittent hemodialysis suits stable patients and rapid clearance (severe hyperkalemia, poisons). Match the modality to hemodynamics and goals.
What CRRT dose should be used?
A delivered effluent dose of about 20–25 mL/kg/h. Because circuit downtime erodes delivery, prescribe roughly 25–30 mL/kg/h to actually deliver 25. Higher doses gave no survival or recovery benefit (RENAL, ATN) and increase hypophosphatemia and drug losses.
What anticoagulation is best for CRRT?
Regional citrate is first-line for most patients: it prolongs filter life span and causes fewer bleeding events than systemic heparin (RICH). Watch for citrate accumulation in severe liver failure or deep shock (a rising total-to-ionized calcium ratio). Use heparin or an anticoagulant-free circuit if citrate is contraindicated.
Do diuretics help acute kidney injury?
No. Loop diuretics do not prevent or treat AKI and do not improve survival or recovery. Use them only to manage fluid overload in a diuretic-responsive patient — never to convert oliguric to non-oliguric AKI or to postpone clearly indicated dialysis.
Can contrast-induced kidney injury be prevented?
The only proven measure is volume expansion with crystalloid in high-risk patients; use the lowest necessary contrast volume. N-acetylcysteine and sodium bicarbonate do not help (PRESERVE). Contrast nephropathy is also less common than once believed in most patients.

16Guidelines & evidence basis

Badges summarize each recommendation’s basis: Guideline: strong Guideline: conditional RCT benefit RCT neutral RCT harm. Guideline strengths use the GRADE classes quoted from the documents below.

Guidelines referenced (2)
  • Kellum JA, Lameire N; KDIGO AKI Guideline Work Group. Diagnosis, evaluation, and management of acute kidney injury: a KDIGO summary (Part 1). Crit Care. 2013;17(1):204. DOI 10.1186/cc11454. PMID 23394211. — the universal AKI definition, staging, and supportive “care bundle”; a KDIGO AKI/AKD update is in public review.
  • Jourdain M, Chatti K, Klouche K, et al. Renal replacement therapy in an intensive care unit: guidelines from the SRLF–GFRUP consensus conference. Ann Intensive Care. 2025;15(1):100. DOI 10.1186/s13613-025-01517-0. PMID 40668437. — 45 statements on RRT indications, modality, dose, monitoring, access, and weaning.

Disclaimer. This page is an educational summary of published evidence for clinicians and is not a substitute for individual clinical judgment or institutional protocols. Definitions, thresholds, doses, and targets are taken from the cited trials and guidelines; verify against current device labeling, local guidelines, and patient-specific factors before applying. “In practice” notes are general practical guidance. Recommendation strengths are quoted from the cited guidelines (GRADE). Reviewed by the DosePilot Medical Team (Jun 2026); all citations verified against PubMed (PMID) and DOI on 2026-06-24.