One-Pager

CRRT (Continuous Renal Replacement)

Modes, dose, circuit pressures, citrate and nursing care for critical care exams

Reviewed by Alex G, BSN, RN, CEN, CCRN, CMC, NREMT-P · Checked against current sources October 2026

Download one-page PDFLetter landscape, ready to print or keep with your notes

Indications: why continuous

CRRT removes fluid and solutes slowly, around the clock, so shifts are small. It suits unstable ICU patients with AKI.

KDIGO suggests CRRT over intermittent HD for hemodynamic instability and for brain injury or high ICP. EASL advises CRRT in acute liver failure, including for persistent high ammonia.

Modalities and dose

All modes can also remove fluid (net ultrafiltration).
ModeMechanismFluid usedClears
SCUFUltrafiltrationNoneFluid only
CVVHConvectionReplacementSmall + middle molecules
CVVHDDiffusionDialysateSmall solutes (K, urea)
CVVHDFBothBothSmall + middle molecules

Typical adult targets for exam study. Follow your institution's protocols.

Access and circuit pressures

PressureUsual signAbnormal means
Access (draw)NegativeMore negative: kink, clot, wall contact, low volume
Filter (pre-filter)Positive, highestRising: clotting or return-side block
ReturnPositiveHigh: kink or clot. Sudden drop: disconnection
TMP (across membrane)PositiveSteady rise: clogging or clotting
Pressure drop (along filter)PositiveRising: clot in the fibers

Anticoagulation

KDIGO suggests regional citrate over heparin when citrate is safe: longer filter life, less bleeding.

FeatureCitrate (regional)Heparin (systemic)
ActionBinds Ca in circuit; Ca replaced; liver makes HCO3Boosts antithrombin
MonitorPost-filter iCa 0.25–0.35; patient iCa 1.1–1.3 mmol/LaPTT or anti-Xa, platelets
RisksLow Ca, alkalosis, accumulationBleeding, HIT
Avoid / cautionSevere liver failure, shock with rising lactate (accumulation)Active bleeding; never in HIT

Low iCa signs: tingling lips or fingertips, cramps, tetany, long QT, seizures.

Metabolic alkalosis usually means citrate overload (citrate becomes bicarbonate). In HIT, stop all heparin, including catheter locks, flushes and priming; use a non-heparin anticoagulant such as argatroban (liver-cleared, aPTT-monitored) per protocol.

Filter clotting

Nursing care: fluid, labs, drugs

Practice CRRT (Continuous Renal Replacement) questions free in your browser27 exam-style questions with rationales. No account or download needed.

In depth

The one-page summary above is what the PDF prints. These sections go further.

Clearance physics and hemodynamics

Diffusion moves solute down a concentration gradient into countercurrent dialysate. At CRRT flows (dialysate 1–2 L/h against blood 9–15 L/h) the dialysate leaves nearly saturated, so clearance of urea (60 Da), creatinine and potassium roughly equals the dialysate rate, while middle molecules (β₂-microglobulin, myoglobin, cytokines) barely diffuse. Convection drags solute across with plasma water pushed by TMP; replacement fluid returns the volume. Adsorption is a small third route (AN69ST, oXiris).

Sieving coefficient = concentration in ultrafiltrate ÷ concentration in plasma water: about 1.0 for urea, electrolytes, glucose and citrate, 0 for albumin (66 kDa) on a 30–40 kDa membrane, roughly the unbound fraction for a drug. Post-dilution convective clearance = ultrafiltration rate × sieving coefficient. Pre-dilution lowers it by about Qb ÷ (Qb + Qpre), typically 15–25%, which is why effluent rate overstates delivered dose (KDIGO 2012).

Filtration fraction = total ultrafiltration (all replacement + net removal) ÷ plasma flow into the filter, with plasma flow = blood flow × (1 − Hct) plus any pre-dilution fluid. Example: Qb 200 mL/min = 12 L/h; Hct 0.30 gives 8.4 L/h plasma; 2 L/h post-dilution plus 0.1 L/h removal = 2.1 L/h; FF = 25%. The same 2 L/h pre-filter gives 2.1 ÷ 10.4 = 20%. Above 20–25% the hemoconcentrated blood leaving the fibers clots (Joannidis 2007).

Once filled, the circuit is a closed loop, so a blood flow of 150–250 mL/min does not itself drop preload; the prime shift, cooling and an AN69 bradykinin reaction cause hypotension at connection. After that, hemodynamics follow net ultrafiltration: 100 mL/h is 2.4 L/day, tolerated only as fast as interstitial fluid refills plasma, which sepsis and hypoalbuminemia slow. Solute shifts are a quarter to a sixth of an intermittent run, so osmotic water movement into brain cells is small, the basis for KDIGO's preference in raised ICP. Correcting acidosis and removing lung water improve pressor response and oxygenation.

Machines, sets and fluids

Baxter Prismaflex and PrisMax share disposable sets, balance fluid on four scales and read four pressures (access, filter, effluent, return); PrisMax adds an integrated TherMax warmer and automatic pod positioning. Fresenius NxStage System One balances fluid volumetrically inside a cartridge (no scales), runs premixed bagged fluid, allows blood flow to 600 mL/min, and uses one fluid line as dialysate (CVVHD) or pre- or post-filter replacement (CVVH); confirm CVVHDF capability on your cartridge. Prismaflex and PrisMax sets are labeled for 72 h of use.

Prismaflex and PrisMax sets; cutoff about 30–40 kDa for all.
SetMembraneArea (m²)Note
M100 / M150AN690.9 / 1.5Bradykinin reaction with ACE inhibitors
ST100 / ST150AN69ST (surface treated)1.0 / 1.5Binds heparin; little bradykinin release
HF1000 / HF1400Polyarylethersulfone1.15 / 1.4No membrane reaction; common adult choice
oXirisAN69ST, heparin grafted1.5Adsorbs endotoxin and cytokines

Fluids are bicarbonate-buffered and differ in potassium (0, 2 or 4 mEq/L), calcium (0 for citrate protocols, 1.25–1.75 mmol/L otherwise), phosphate (0, or 1.0–1.25 mmol/L in Phoxillum) and glucose (0 or 100 mg/dL). Most bags run as dialysate or replacement, but replacement enters the blood, so its label must allow that. Mix two-chamber bags before hanging. Lactate-buffered fluid worsens acidosis in shock and liver failure.

Patient selection and timing

AEIOU covers the emergent starts (KDIGO 2012). Most ICU starts are for a trajectory: oliguric AKI with a positive balance the lungs cannot take, or an obligatory load of feeds, antibiotics and blood products the kidneys cannot clear. Five trials show that starting on a stage 2 or 3 label without an indication does not save lives, and about 40% of patients managed expectantly never need RRT. The working rule from STARRT-AKI and AKIKI-2: wait for an indication or AKI persisting past 72 h, but not past a BUN of about 140 mg/dL or 72 h of oliguria.

Choose CRRT over intermittent HD for pressor dependence, brain injury or raised ICP (both 2B in KDIGO 2012), acute liver failure with hyperammonemia (EASL 2017), or overload beyond what 4 h of HD can remove. Choose intermittent HD or SLED when a toxin needs fast clearance (lithium, methanol, metformin, salicylate per EXTRIP), potassium must fall within the hour, or the patient is stable. CRRT has no absolute contraindication of its own; the limits are goals of care, no safe access, and bleeding risk, which changes the anticoagulant rather than the decision.

The prescription, initiation and fluid accounting

A complete order states modality, set, anticoagulant, blood flow, dialysate rate, replacement split pre and post filter, net ultrafiltration with an hourly cap and a 24 h goal, each line's fluid with its potassium, calcium and phosphate content, warmer setting, lab schedule and dose target. Dose = total effluent (dialysate + replacement + net removal + citrate pre-blood pump fluid) ÷ weight. Deliver 20–25 mL/kg/h (KDIGO 2012, 1A); prescribe 25–30 to cover downtime and pre-dilution.

Worked CVVHDF order for an 80 kg adult.
SettingValueReason
Blood flow200 mL/min (100–150 on citrate)Keeps FF low; limits citrate load
Dialysate1,000 mL/hSmall-solute clearance
Replacement pre / post500 / 500 mL/hPre lowers FF; post gives full clearance
Net removal0–150 mL/h, reset hourlyFollows the balance goal
Effluent2,100–2,150 mL/h ≈ 27 mL/kg/hDelivers 20–25 after downtime
FluidK 4 mEq/L (2 if K above 5), phosphate 1.0Prevents depletion

Before connection: both lumens draw and flush freely; withdraw each lumen's lock (volume printed on the hub) rather than flushing it in; baseline labs drawn; every hung bag matched to the order. The 150–200 mL saline prime goes to waste for an overloaded patient and to the patient when hypotensive. Start blood flow at 100 mL/min and step up over 10–15 minutes; leave net removal at zero for the first 30–60 minutes.

The machine removes exactly what it is set to; the patient's balance is that minus everything else. Each hour, total intake (every drip, feed, flush, blood product, citrate and calcium) and output (urine, drains, stool, NG), then set the next hour's removal so the running total moves toward the goal set on rounds (Prowle and Mehta 2021). Hold or cut removal for a falling MAP or rising pressor. Rates above about 1.75 mL/kg/h carried higher mortality in RENAL data (Murugan 2019): a 70 kg patient rarely needs more than 120 mL/h outside pulmonary edema.

Catheter, circuit pressures and alarms

Adults need a 12–14 Fr double-lumen catheter reaching the cavoatrial junction: about 15 cm right IJ, 20 cm left IJ, 24–25 cm femoral (tip in the IVC). The lock (heparin 1,000–5,000 units/mL or 4% citrate, to the printed lumen volume) is withdrawn before use; chlorhexidine dressing changed every 7 days or when soiled. Catheter flow problems cause most clotting (Tandukar 2019): access pressure more negative than about −150 mmHg, or blood flow that cannot hold 150 mL/min, means the catheter, not the filter.

  1. Clamps, kinks, dressing tension, lines under the patient
  2. Reposition the patient (head turn, leg straight, roll); drop blood flow 50 mL/min
  3. Flush each lumen with 10 mL saline and aspirate; alteplase 1–2 mg dwell for a dead lumen per protocol
  4. Reverse the lines and document it; recirculation rises toward 20% and dose falls
  5. X-ray for tip position; call for exchange if flow still fails
Prismaflex and PrisMax; typical adult values, confirm on your machines.
PressureTypicalAlarmChange means
Access−50 to −150 mmHg−250 extremeMore negative: catheter, kink, clamp, low volume
Filter+100 to +250Through TMP and ΔPHighest in the circuit; rises as fibers clot
Return+50 to +150+350; below +10 disconnectionHigh: return kink or clot. Drop: disconnection
TMP20–150, slow rise300 excessiveFouling or clotting; sudden rise: clot

On Prismaflex and PrisMax, TMP ≈ (filter + return) ÷ 2 − effluent and pressure drop = filter − return; both rise before a filter fails (Hu 2026), so a steady TMP climb is the cue to change the set on your terms with the blood returned. NxStage System One reads arterial, venous and effluent pressure through pods and sets alarm limits around the running pressure; it shows no Prismaflex-style pressure drop, so a climbing venous pressure with a stable arterial pressure is its clotting signal.

AlarmLikely causeCheck and act
Access extremely negativeCatheter on the vein wall, kink, clamp, hypovolemia, coughClamps and kinks; reposition; lower Qb; flush; reverse lines
Return extremely positiveKink or clamp after the filter; clot in chamber or lumenUnclamp; check chamber; flush return lumen; reverse lines
Air in bloodLoose access-side connection; low chamber levelStop, clamp, tighten, raise chamber level; never override
TMP excessive, filter clottingFouling, FF too high, weak anticoagulationCut net UF, shift to pre-dilution, raise Qb; plan change
Filter clottedFibers thrombosedNever force blood back; new set; count about 150 mL lost
Blood leakFiber rupture; false positive from bilirubin or lipidDipstick effluent; positive: new set; negative: reset
Scale or bag alarmsBag touching, swinging, wrong scale, spike closed, emptyHang bags free; change fast; 5 L lasts 2 h at 2.5 L/h

Every stopped minute raises clot risk; past about 10 minutes expect to lose the circuit. Call the provider or CRRT team for a catheter that fails the flow steps, repeated early clotting (think HIT or filtration fraction), bleeding, any anticoagulant change, and the decision to return blood and stop. Call rapid response for suspected air embolism (clamp, left lateral, 100% O₂) or a disconnection with large blood loss.

Anticoagulation: citrate, heparin, none

Citrate is regional: infused pre-filter (the pre-blood pump line on Prismaflex) at about 3–4 mmol per liter of blood flow, it chelates ionized calcium below the 0.35 mmol/L clotting needs. Prismocitrate 18/0 (18 mmol/L) at 1,500–2,500 mL/h suits blood flows of 100–150 mL/min; 4% trisodium citrate (136 mmol/L) and ACD-A (about 113 mmol/L) need far less volume but add sodium. A third to a half of the citrate-calcium complex leaves in the effluent; the rest reaches the patient, where liver, muscle and kidney turn each citrate into three bicarbonate within minutes. Fluids are calcium-free, and calcium chloride or gluconate runs into the return line or a separate central line to hold patient iCa at 1.0–1.3 mmol/L.

MeasureTargetWhenOff target
Post-filter iCa0.25–0.35 mmol/L (KDIGO 0.2–0.4)1 h after start or change, then q6hHigh: more citrate; low: less
Patient iCa1.0–1.3 mmol/L1 h, then q6hLow: more calcium; tetany, long QT
Total Ca ÷ iCaUnder 2.5 (same units; mg/dL ÷ 4)q12–24h2.5 or more: accumulation
pH, HCO₃, anion gapNormalq6hAlkalosis: overload; gap acidosis: accumulation

Accumulation means the patient cannot metabolize citrate (liver failure, shock with lactate above about 4 mmol/L, hypothermia): patient iCa falls while the calcium infusion climbs, total calcium rises, the ratio passes 2.5 and the anion gap widens (2026 Delphi consensus). Cut citrate by 25–50% or stop it, raise dialysate flow to remove more, and switch to heparin or none. Alkalosis without accumulation is overload: lower the citrate dose or raise dialysate flow. In RICH (JAMA 2020) citrate beat heparin on filter life and bleeding at the cost of more infections, with no mortality difference (evidence table); KDIGO 2012 suggests citrate first (2B) and the 2026 draft grades it 1B.

Unfractionated heparin is systemic: bolus 500–1,000 units, then 5–10 units/kg/h into the pre-filter port, titrated to aPTT about 1.5× control or anti-Xa 0.3–0.4 IU/mL; filter life 20–30 h. A platelet fall of 50% or more from day 5–10 with early, repeated clotting is HIT until proven otherwise: stop all heparin including locks and primes, send PF4 antibody, start argatroban 0.2–2 mcg/kg/min (lower in liver failure). No anticoagulant suits active bleeding or coagulopathic liver failure: blood flow 200–250 mL/min, CVVHD or pre-dilution to keep FF under 20%, saline flushes of 100–200 mL every 1–2 h counted as intake, filter life 12–24 h.

Complications: drugs, nutrition, electrolytes, heat

CRRT at 25 mL/kg/h behaves like a glomerular filtration rate of 25–35 mL/min that never pauses, so renally cleared drugs are dosed as for a creatinine clearance of 25–50 mL/min, and loading doses are not reduced. A drug is removed when it is small (nearly all are), mostly unbound and has a small volume of distribution. Removed: beta-lactams, vancomycin, aminoglycosides, fluconazole, levetiracetam, lithium, metformin. Not removed: amiodarone, digoxin, phenytoin, propofol, echinocandins, warfarin. Underdosing antibiotics is the common error; the second is leaving CRRT doses in place after the circuit stops. Pharmacist review daily and at every change in effluent rate or downtime.

Effluent carries 10–15 g of amino acids a day plus water-soluble vitamins, trace elements and glucose if the fluid has none. Protein targets differ: KDIGO 2012 caps at 1.7 g/kg/day, ASPEN/SCCM 2016 allows up to 2.5; most units feed 1.5–2.0 g/kg/day at 25–30 kcal/kg and count citrate's few hundred kcal/day. Hypophosphatemia appears in most patients on phosphate-free fluid by day 2–3 and weakens the diaphragm; phosphate-containing fluid prevents it. Potassium follows the bag (4 mEq/L once K is under 5), and magnesium needs replacing on citrate because it is chelated too.

Blood cools in the lines and fluids enter at room temperature; without a warmer core temperature drops 1–2 °C, masking fever, raising SVR and impairing coagulation. Set the warmer to 37–38 °C; integrated warmers cut the risk without removing it (Bell 2024). Other complications and first steps: bleeding on heparin (hold, move to citrate); catheter infection (remove if suspected); AN69 membrane reaction (hypotension minutes after start, worse on ACE inhibitors: stop, change set); and over-removal, which shows as a rising pressor dose and delays renal recovery.

Daily care, documentation, transport and liberation

Document hourly: pressures with TMP and ΔP, blood and fluid rates, removal set and achieved, effluent volume, running balance, fluid additives, anticoagulant dose and labs, warmer temperature, every alarm with its fix, and every minute of downtime, since downtime is the gap between prescribed and delivered dose. Machines carry a battery for a power interruption, not a trip: for CT or the OR either return the blood and disconnect (expect to lose the set if the stop exceeds 10–15 minutes) or move with the machine and a second trained nurse.

Recovery is read from urine: spontaneous output above about 400–500 mL/day without diuretic, rising day over day, and a timed 2–6 h creatinine clearance of 15–20 mL/min or more; a furosemide stress test (1–1.5 mg/kg, urine above 200 mL in 2 h) predicts response. KDIGO 2012 suggests not using diuretics to shorten RRT (2B). Move to intermittent HD or SLED when MAP holds on little or no pressor and 1–2 L can come off in 4 h. After stopping: BMP and potassium at 6–12 h and 24 h, re-cut renally cleared drugs, and watch 48 h for the indication to return. Nephrology follow-up at discharge.

Evidence and controversies

Timing of initiation.
TrialComparisonResult
ELAIN 2016 (n 231, surgical)Within 8 h of stage 2 vs 12 h of stage 390-day mortality 39% vs 55%, favoring early
AKIKI 2016 (n 620)At stage 3 vs wait for an indication60-day mortality 48.5% vs 49.7%; 49% never dialyzed
IDEAL-ICU 2018 (n 488, septic shock)Within 12 h vs 48 h of failure-stage AKI90-day mortality 58% vs 54%; stopped for futility
STARRT-AKI 2020 (n 3,019)Within 12 h of stage 2–3 vs indication or 72 h90-day mortality 43.9% vs 43.7%; dependence 10.4% vs 6%
AKIKI-2 2021 (n 278)Indication or 72 h oliguria vs BUN over 140No gain from waiting longer; more deaths with it
Dose, modality and anticoagulation.
TrialComparisonResult
ATN 2008 (n 1,124)35 vs 20 mL/kg/h60-day mortality 53.6% vs 51.5%
RENAL 2009 (n 1,508)CVVHDF 40 vs 25 mL/kg/h90-day mortality 44.7% in both
IVOIRE 2013 (n 140, septic shock)70 vs 35 mL/kg/h28-day mortality 37.9% vs 40.8%
Hemodiafe 2006 (n 360)Intermittent HD vs CVVHDF60-day survival 32% vs 33%
RICH 2020 (n 638)Citrate vs heparinFilter 47 vs 26 h; bleeding 5% vs 17%; mortality same

What changed practice: a delivered dose of 20–25 mL/kg/h (KDIGO 2012, 1A) and no high-volume hemofiltration for sepsis; expectant timing with firm stop-rules rather than starting on a stage label. Still open: the right trigger inside the expectant window (AKIKI-2 says not past BUN 140 mg/dL); whether CRRT beats intermittent HD for anyone (trials are neutral; the preference rests on physiology); the dialysis-dependence signal after early starts; and fluid-removal limits, which rest on observational data. KDIGO posted a public-review draft of the 2026 AKI and AKD guideline in March 2026; the final version was still in preparation in October 2026, so KDIGO 2012 remains the citable guideline and the draft's grades (dose 1B, citrate over heparin 1B, deferred start 1C) are provisional.

Exam pearls

Practice free in your browser3,800+ exam-style questions with rationales. No account, no ads. Also free on the App Store and Google Play.

Related one-pagers

Calculators for this topic

Sources

  1. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO Clinical Practice Guideline for Acute Kidney Injury. Kidney Int Suppl. 2012;2:1-138. Link
  2. KDIGO. Acute Kidney Injury (AKI) and Acute Kidney Disease (AKD) guideline page (status of the 2026 update). Accessed October 5, 2026. Link
  3. European Association for the Study of the Liver. EASL Clinical Practical Guidelines on the management of acute (fulminant) liver failure. J Hepatol. 2017;66:1047-1081. Link
  4. Tandukar S, Palevsky PM. Continuous Renal Replacement Therapy: Who, When, Why, and How. Chest. 2019;155(3):626-638. Link
  5. Saunders H, Rehan A, Hashmi MF, Sanghavi DK. Continuous Renal Replacement Therapy. StatPearls. Updated March 10, 2024. Link
  6. Anastasopoulou C, Goyal A. Hypocalcemia. StatPearls. Updated June 8, 2026. Link
  7. Jacobs R, et al. Metabolic complications of citrate anticoagulation in continuous renal replacement therapy (CRRT): a Delphi consensus on indications, monitoring and management. Crit Care. 2026;30:358. Link
  8. Higgins C. Citrate anticoagulation during CRRT for acute kidney injury. acutecaretesting.org. April 2017. Link
  9. Bradford Teaching Hospitals NHS Foundation Trust ICU. Protocol for Continuous Renal Replacement Therapy using Regional Citrate, v2.1. December 2017 (hosted by CRRT Online). Link
  10. Karakala N. Alarms and troubleshooting (CRRT 2017 conference presentation). CRRT Online. Link
  11. PrismaFlex System (Baxter) pressure sensor range, accuracy and alarm limits. Supplemental Digital Content 1, ASAIO Journal, 2022. Link
  12. Kakajiwala A, et al. Membrane pressures predict clotting of pediatric continuous renal replacement therapy circuits. Pediatr Nephrol. 2017;32(7):1251-1261. Link
  13. Joannidis M, Oudemans-van Straaten HM. Clinical review: Patency of the circuit in continuous renal replacement therapy. Crit Care. 2007;11:218. Link
  14. Hu Y, et al. Prediction of filter clotting using longitudinal trends of circuit pressure parameters during continuous renal replacement therapy (CRRT): an exploratory study. Sci Rep. 2026;16:17411. Link
  15. Prowle J, Mehta R. Fluid balance management during continuous renal replacement therapy. Semin Dial. 2021;34:440-448. Link
  16. Murugan R, 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. Link
  17. Bell M, et al. Modern CRRT systems are associated with lower risk of hypothermia. Sci Rep. 2024;14:23162. Link
  18. Pornsirirat T, et al. Incidence of hypothermia in critically ill patients receiving continuous renal replacement therapy in Siriraj Hospital, Thailand. Acute Crit Care. 2024. Link
  19. EXTRIP Workgroup. Lithium: extracorporeal treatment recommendations. Link
  20. EXTRIP Workgroup. Methanol: extracorporeal treatment recommendations. Link