Renal Replacement Therapy Modalities — IHD vs CVVH vs CVVHDF vs SLED

Overview & definition

The renal replacement therapy (RRT) removes the solutes (the urea, the creatinine, the potassium, the toxins) and the fluid (the ultrafiltration) when the kidneys fail. The four modalities differ in the solute-transport mechanism (the diffusion vs the convection), the duration and the intermittency, and the haemodynamic tolerance. The choice depends on the haemodynamic stability, the solute-removal goal, the fluid-balance need, and the bleeding risk.[1]

The solute-transport mechanisms

• The diffusion — the solute moves down its concentration gradient across the semipermeable membrane (the dialysate flows counter-current to the blood). Efficient for the small solutes (the urea, the creatinine, the potassium). The basis of the haemodialysis (the IHD).[1]
• The convection — the solute is dragged with the water across the membrane (the solvent drag) by the hydrostatic pressure; the bulk flow removes the small AND the middle molecules. The basis of the haemofiltration (the CVVH).[1]
• The ultrafiltration — the fluid (water) removal by the hydrostatic pressure (the plasma water dragged across the membrane). The fluid-balance control.[1]

The modalities

[1]

Intermittent haemodialysis (IHD)

• The diffusion-based, the intermittent, the 3 to 4 hours a session, three times a week (or daily in the AKI). The high blood and the dialysate flows.[1]
• The rapid solute clearance — the efficient small-solute removal.[1]
• The rapid fluid and solute shifts — the hypotension (the risk for the haemodynamically unstable), the dialysis disequilibrium (the rapid solute shifts → the cerebral oedema, the risk in the uraemia and the raised ICP).[1]
• For the stable patient (the stable AKI, the CKD, the end-stage).[1]

Continuous veno-venous haemofiltration (CVVH)

• The convection-based, the continuous (the 24 hours), the pumped veno-venous access (a dual-lumen central catheter).[1]
• The middle-molecule clearance (the larger solutes, the inflammatory mediators).[1]
• The gentle, the haemodynamically tolerated — the slow, the continuous fluid removal avoids the rapid shifts. The good for the haemodynamically unstable, the catabolic, and the raised ICP.[1]
• Needs the anticoagulation (the regional citrate or the heparin).[1]

Continuous veno-venous haemodiafiltration (CVVHDF)

• The combined diffusion and convection — the dialysate (the diffusion) plus the replacement fluid (the convection).[1]
• The combined clearance — better than the CVVH alone. The commonest CRRT modality.[1]
• The continuous, the gentle, the haemodynamically tolerated.[1]

Sustained low-efficiency dialysis (SLED)

• The hybrid — a slow, the extended (the 6 to 10 hours) intermittent haemodialysis, the daily.[1]
• A bridge between the IHD and the CRRT — the better haemodynamic tolerance than the IHD, the less anticoagulation than the CRRT.[1]
• Useful when the CRRT is unavailable or for the transitional phase.[1]

The indications — the AEIOU

The RRT is indicated for:[1]

• A — the acidosis refractory to the bicarbonate.[1]
• E — the electrolyte disorders refractory to the medical therapy (the severe or the refractory hyperkalaemia).[1]
• I — the intoxications with the dialysable toxins (the lithium, the salicylate, the metformin, the methanol, the ethylene glycol).[1]
• O — the overload — the refractory fluid overload (the pulmonary oedema).[1]
• U — the uraemia — the complications (the uraemic pericarditis, the uraemic encephalopathy, the uraemic bleeding).[1]

Plus the oliguria with the inability to meet the fluid and the nutritional needs.[1]

The timing — the early vs the late

The AKIKI, the STARRT-AKIN, and the ELAIN trials addressed the early vs the late RRT in the moderate-to-severe AKI. The consensus: no clear benefit of the early RRT in the moderate AKI without the absolute indication — the standard is to wait for the absolute indication (the refractory K, the acidosis, the overload, the uraemia) unless the uraemic complications or the fluid overload develop. The early RRT increases the exposure to the dialysis (the catheter, the anticoagulation) without the outcome benefit. The ELAIN trial suggested a benefit of the earlier (the furosemide-staged) RRT, but this is controversial and not the standard.[1]

The anticoagulation

• The regional citrate — the preferred. The citrate chelates the calcium in the circuit (the local anticoagulation) and is metabolized in the liver (the systemic anticoagulation avoided). The monitoring: the post-filter ionized calcium (low), the systemic ionized calcium (normal), the total calcium rising. The risks: the metabolic alkalosis (the citrate metabolized to the bicarbonate), the hypercalcaemia, and the citrate accumulation (the total-to-ionized calcium ratio over 2.5) in the severe liver failure.[1]
• The unfractionated heparin — the effective, the cheap; the bleeding risk and the heparin-induced thrombocytopenia (the HIT).[1]
• The low-molecular-weight heparin, the prostacyclin.[1]
• The no anticoagulation — for the high-bleeding-risk patient; the saline flushes; the shorter circuit life.[1]

The dose

The CRRT dose (the effluent rate) is the 20 to 25 mL/kg/h. The ATN and the RENAL trials showed that the higher dose (the 35 mL/kg/h) offered no benefit over the 20 to 25 mL/kg/h. The delivered dose is less than the prescribed (the downtime for the filter change, the investigations, the clotting).[1]

[1] [1] [1]

Red flags

The modality selection — matching the patient to the machine

The choice of the modality is dictated by three questions: (1) Is the patient haemodynamically stable enough to tolerate the rapid solute and fluid shifts of the IHD? (2) Is the dominant goal the small-solute clearance (the urea, the potassium — the IHD or the CVVHD), the middle-molecule and the mediator clearance (the CVVH), or the balanced removal (the CVVHDF)? (3) Is the resource available (the CRRT machine, the trained nursing, the citrate protocol, the 24-hour commitment)? The KDIGO guideline and the CICM consensus accept that no modality has shown a clear survival advantage in the unselected AKI population — the choice is therefore a physiological and a logistical one, not an evidence-mandated one.[1][1][1]

The general rule: the CRRT (the CVVHDF) for the haemodynamically unstable, the catabolic, the raised intracranial pressure, and the patient who needs the precise fluid and electrolyte control; the IHD for the stable, the recovering, and the patient on a chronic programme; the SLED as the hybrid bridge when the CRRT is unavailable or the IHD is poorly tolerated. The SLED and the CRRT show comparable haemodynamic tolerance and the solute control; the SLED is the pragmatic compromise.[1][1]

[1]

Intermittent haemodialysis — the deep dive

The IHD is the diffusion-based, the high-efficiency, the intermittent modality. The blood and the dialysate flow counter-current at high rates (the blood 250 to 400 mL/min, the dialysate 500 to 800 mL/min) through a high-flux dialyser; the urea, the creatinine, and the potassium move down their gradients into the dialysate and are discarded.[1][1]

The defining physiological feature is the rapid solute clearance and the rapid fluid shift. A three-to-four-hour session removes the solute and the fluid load of one to two days, which produces: (a) the osmotic shift — the rapid fall of the plasma urea lowers the plasma osmolality, the water moves into the cells, and the dialysis disequilibrium (the headache, the nausea, the seizures, the cerebral oedema) follows; and (b) the volume shift — the rapid ultrafiltration lowers the plasma volume faster than the refilling from the interstitial space, and the intradialytic hypotension follows. Both are the price of the efficiency.[1]

The dose is measured by the single-pool Kt/V (the target 1.2 per session in AKI, the delivered 1.3 to compensate for the rebound) or the urea reduction ratio (the target 65 per cent). The daily IHD is used in the catabolic AKI when the alternate-day dosing does not control the uraemia or the potassium.[1]

The advantages — the rapid correction of the hyperkalaemia and the acidosis, the short session time, the lower cost, the patient mobility between sessions, the established chronic infrastructure. The disadvantages — the haemodynamic intolerance (the intradialytic hypotension in the vasopressor-dependent patient), the dialysis disequilibrium in the markedly uraemic patient, the rebound hyperkalaemia between sessions, and the intermittent nature that does not allow the precise fluid and electrolyte control.[1][1]

The dialysis disequilibrium — the prevention

The dialysis disequilibrium is prevented by the slow first dialysis (the reduced blood flow, the reduced time, the two-hour session in the markedly uraemic — the urea over 35 mmol/L), the osmotic protection (the mannitol or the hypertonic saline if the disequilibrium threatens), and the avoidance in the raised intracranial pressure (the CRRT preferred in the head injury and the hepatic failure). The disequilibrium is rare in the chronic patient (the slow uraemia) and dangerous in the acute (the rapid rise).[1]

Continuous renal replacement therapy — the deep dive

The CRRT is the slow, the continuous, the pumped-veno-venous family of modalities. The blood is drawn from a central dual-lumen catheter, driven by a peristaltic pump through a high-flux haemofilter, and returned to the patient; the fluid (the effluent) is removed and the replacement (the substitution) fluid is added to maintain the balance. The CRRT runs 24 hours a day, the solute and the fluid removal is the gradual, the haemodynamic tolerance is the excellent, and the fluid and the electrolyte control is the precise.[1][1]

The three CRRT sub-modalities differ in the solute-transport mechanism and therefore in what they clear best: [1]

Continuous veno-venous haemofiltration (CVVH)

The pure convection. The hydrostatic pressure drives the plasma water (the solvent drag) across the high-flux membrane; the solutes are dragged with the water in proportion to their sieving coefficient. The effluent is the ultrafiltrate; the lost volume is replaced by the substitution fluid (the pre- or the post-dilution). The CVVH clears the middle molecules (the beta-2-microglobulin, the cytokines, the inflammatory mediators) better than the pure diffusion — the rationale for the CVVH in the sepsis and the systemic inflammation. The clearance is set by the ultrafiltration (the effluent) rate.[1]

Continuous veno-venous haemodialysis (CVVHD)

The pure diffusion. The dialysate flows counter-current to the blood through the filter; the solutes move down their gradient. No substitution fluid is used (no convective component), so the middle-molecule clearance is the limited and the small-molecule clearance dominates. The CVVHD is the simplest, the cheapest (no replacement fluid), and the easiest to anticoagulate (the lower filtration fraction). The clearance is set by the dialysate flow rate.[1]

Continuous veno-venous haemodiafiltration (CVVHDF)

The combined diffusion and convection — the dialysate (the diffusion) plus the substitution fluid (the convection). The combined clearance is the additive; the CVVHDF delivers the small- and the middle-molecule clearance simultaneously and is the commonest CRRT modality worldwide (the RENAL and the ATN trials both used the CVVHDF). The effluent is the sum of the dialysate and the ultrafiltrate, and the dose (the 20 to 25 mL/kg/h) is set by the total effluent rate.[1][1]

[1]

The pre-dilution vs the post-dilution

The substitution fluid in the CVVH and the CVVHDF may be delivered before the filter (the pre-dilution) or after the filter (the post-dilution). The post-dilution gives the higher clearance (the full concentration of the solute at the membrane) but the higher filtration fraction (the haemoconcentration in the filter, the clotting risk). The pre-dilution dilutes the blood before the filter (the lower filtration fraction, the longer filter life) at the price of the 15 to 20 per cent lower clearance. The compromise — the 1:2 pre-to-post split — balances the filter life and the clearance. The pure pre-dilution is the rescue for the high haematocrit and the high filtration fraction.[1]

Sustained low-efficiency dialysis — the deep dive

The SLED (and its variants — the sustained low-efficiency daily diafiltration, the go-slow IHD) is the hybrid that uses the standard IHD machine at the reduced blood and dialysate flows (the blood 150 to 200 mL/min, the dialysate 100 to 300 mL/min) for the extended duration of the 6 to 12 hours. The reduced efficiency and the extended time deliver the solute and the fluid removal at the slower rate — the haemodynamic tolerance approaching the CRRT, the anticoagulation need reduced (the shorter circuit time than the CRRT), and the cost and the infrastructure closer to the IHD.[1][1]

The SLED is the answer when: the patient is too unstable for the standard IHD but the CRRT machine is unavailable; the unit runs the CRRT only by the day (the SLED overnight); the patient is the transitional phase (the weaning from the CRRT to the IHD); or the resource setting cannot sustain the 24-hour CRRT. The SLED and the CRRT show the comparable mortality and the renal recovery in the head-to-head trials, and the SLED uses the less anticoagulation and the less nursing time.[1][1]

The membrane and the dialyser

The membrane is the semipermeable barrier across which the diffusion and the convection occur. The modern ICU membranes are the high-flux, the biocompatible, the synthetic (the polysulfone, the polyethersulfone, the polyamide, the polyacrylonitrile [the AN69]). The high-flux membrane (the ultrafiltration coefficient KUF over 20 mL/h/mmHg/m²) allows the convective removal of the middle molecules; the biocompatible synthetic membrane (the low complement and the low leukocyte activation) is preferred over the bioincompatible cellulose-based membrane (the cuprophane) — the biocompatible membrane reduces the inflammation, the delayed renal recovery, and possibly the mortality in the AKI.[1][1]

The sieving coefficient (the SC) is the ratio of the solute concentration in the ultrafiltrate to the plasma; the SC of 1.0 (the urea, the creatinine, the potassium, the sodium) means the free passage, and the SC of 0 (the albumin, the larger proteins) means the complete retention. The middle molecules (the beta-2-microglobulin, the vitamin B12) have the intermediate SC and are cleared by the convection. The drug removal depends on the SC, the protein binding, and the volume of distribution.[1][1]

The AN69 membrane has the particular property of the bradykinin-release syndrome (the BRS) — the contact of the negatively charged AN69 surface with the blood (especially in the patient on the ACE inhibitor) triggers the bradykinin generation and the anaphylactoid reaction. The AN69-ST (the surface-treated, the heparin-bonded) and the polymethylmethacrylate membranes reduce this risk. The BRS presents as the hypotension and the flushing in the first minutes of the circuit and is managed by the pre-rinsing, the stopping the ACE inhibitor, and the membrane change.[1]

The dialysate and the replacement fluid

The dialysate (the IHD and the CVVHD/CVVHDF) and the replacement (the substitution) fluid (the CVVH and the CVVHDF) carry the electrolytes that diffuse into the patient. The standard bicarbonate-buffered fluid contains the sodium 140, the potassium 2 to 4 (the adjustable), the calcium 1.5 to 1.75, the magnesium 0.5 to 1.0, the chloride 100 to 115, and the bicarbonate (or the lactate, metabolised to the bicarbonate) 30 to 35 mmol/L. The bicarbonate buffer is preferred over the lactate (the bicarbonate avoids the lactate load in the lactic acidosis and the liver failure).[1][1]

The phosphate is absent (or very low) in the standard fluid, so the prolonged CRRT routinely produces the hypophosphataemia (the respiratory muscle weakness, the weaning failure, the rhabdomyolysis) — the phosphate is replaced systemically or via a phosphate-containing replacement fluid. The glucose in the fluid (the 1 to 4 g/L depending on the bag) contributes the glucose load and must be counted in the glycaemic control. The potassium in the fluid is set by the patient's potassium (the low-potassium bag for the hyperkalaemia).[1]

The vascular access

The temporary (the non-tunnelled) dual-lumen central catheter is the standard ICU access. The right internal jugular is the preferred site (the straight course, the lower stenosis, the lower pneumothorax); the femoral is the alternative (the higher infection and the stenosis, but the acceptable in the coagulopathy and the obese); the subclavian is avoided in the AKI (the stenosis risk that precludes the future arteriovenous fistula). The KDIGO and the CDC recommend the ultrasound guidance, the maximal sterile barrier, and the chlorhexidine dressing.[1][1]

The catheter dysfunction (the poor flow, the recirculation) is the commonest access problem and is managed by the repositioning, the urokinase or the alteplase lock, and the exchange over a wire if the persistent. The catheter-related bloodstream infection (the CRBSI) is the major morbidity — prevented by the aseptic insertion, the chlorhexidine patch, the timely removal, and the avoidance of the femoral site in the high-risk.[1]

Citrate anticoagulation — the deep dive

The regional citrate is the default anticoagulant for the CRRT. The trisodium citrate is infused into the blood before the filter; it chelates the ionised calcium (the iCa) — the fourth factor in the coagulation cascade — and the calcium-depleted blood will not clot in the circuit. The calcium-depleted blood (the low post-filter iCa, the 0.25 to 0.40 mmol/L) returns to the patient; the systemic calcium is repleted by the separate calcium chloride or calcium gluconate infusion into the return line, restoring the patient's iCa to the normal 1.10 to 1.30 mmol/L. The result — the circuit is anticoagulated, the patient is not.[1][1]

The citrate is metabolised by the liver (and, to the lesser extent, the skeletal muscle and the renal cortex) to the bicarbonate — one molecule of the citrate yields three molecules of the bicarbonate. This metabolism is the source of the two characteristic derangements: the metabolic alkalosis (the excess bicarbonate) and, when the metabolism fails (the severe liver failure, the severe lactic acidosis, the profound shock), the citrate accumulation.[1]

The citrate accumulation is recognised by the rising total calcium with the stable or falling ionised calcium — the CaT/CaI ratio over 2.5 is the sentinel. The accumulating citrate chelates the calcium systemically, the measured total calcium rises (the citrate-bound calcium is measured but not active), the ionised calcium falls, and the unmetabolised citrate adds the unmeasured anion, raising the anion gap and worsening the metabolic acidosis. The management — the reduce or the stop the citrate, the switch to the heparin or the no-anticoagulation, the do-not-rechallenge until the hepatic function and the lactate recover. The nafamostat is the alternative in the persistent hepatic failure.[1]

The metabolic alkalosis from the citrate is common and is managed by the reduced citrate, the increased acid generation (the chloride load), or the dialysate with the lower bicarbonate. The hypercalcaemia is occasional (the calcium repletion). The hypomagnesaemia and the hypokalaemia may accompany the alkalosis.[1]

[1]

Drug dosing on the CRRT

The drug dosing on the CRRT is the challenge — the continuous clearance removes the drugs that the failing kidney would have retained, and the standard renal-failure dosing (which assumes the minimal clearance) underdoses the patient. The principles:[1][1]

1. The loading dose is unchanged — the loading dose depends on the volume of distribution, not the clearance. Give the full loading dose regardless of the CRRT.[1]
2. The maintenance dose depends on the clearance — the CRRT adds the clearance equal to the effluent rate (the 20 to 25 mL/kg/h, roughly the 1.5 L/h in the 70-kg patient, or the 25 to 35 mL/min). The drugs cleared by the CRRT (the low protein binding, the small Vd, the water-soluble — the beta-lactams, the aminoglycosides, the vancomycin, the linezolid is borderline) need the higher maintenance dose than the standard renal-failure chart. The drugs NOT cleared by the CRRT (the high protein binding, the large Vd, the lipid-soluble — the macrolides, the fluoroquinolones partly, the amphotericin) need the standard or the reduced dose.[1]
3. The therapeutic drug monitoring drives the dose — the vancomycin (the trough 15 to 20 mg/L), the aminoglycosides, the beta-lactams (the trough over the MIC, the 4× the MIC for the time-dependent) are monitored and the dose is adjusted. The CRRT variability (the filter downtime, the flow changes) makes the TDM essential.[1][1]
4. The effluent rate drives the clearance — the higher the effluent, the higher the clearance; the change of the dose (the 20 to 35 mL/kg/h) changes the drug clearance. The citrate does not directly affect the drug clearance but the fluid composition may affect the protein binding.[1]
5. The anticoagulants, the antiepileptics, the antibiotics are the high-stakes drugs — the vancomycin 1 g every 12 to 24 h (the TDM), the meropenem 1 g every 8 h (the beta-lactam), the piperacillin-tazobactam 4.5 g every 6 to 8 h (the beta-lactam), the levetiracetam 500 to 1000 mg every 12 h. The aminoglycosides are cleared partially and need the extended interval with the TDM.[1]

The complications of the RRT

[1]

The intradialytic hypotension (the IHD) is the commonest haemodynamic complication — the rapid ultrafiltration drops the intravascular volume faster than the refilling. The prevention — the reduced UF rate (the longer session, the daily treatment), the lower dialysate temperature (the 35 to 36 °C — the cool dialysate stabilises the haemodynamics), the sodium profiling, the avoid the antihypertensives, the midodrine in the recurrent. The refractory hypotension is the indication to switch to the CRRT.[1]

The filter clotting is the commonest mechanical complication of the CRRT — the cause is the inadequate anticoagulation (the under-dosed citrate or heparin), the high filtration fraction (the over 25 per cent — the haemoconcentration in the filter), the low blood flow (the stasis), the air in the circuit, or the vascular access kinking. The prevention — the adequate anticoagulation (the post-filter iCa 0.25 to 0.40), the filtration fraction under 25 per cent (the add the pre-dilution), the blood flow over 150 mL/min, the de-airing. The clotting loses the dose (the downtime) and the blood (the circuit volume).[1]

The hypothermia of the CRRT is the under-recognised — the extracorporeal circuit loses the heat, the patient cools, and the febrile septic patient may mask the fever on the CRRT (the CRRT cools the patient). The warming of the fluids (the in-line warmer) and the active warming restore the temperature. The hypothermia may mask the fever and delay the infection recognition.[1]

The weaning and the transition

The RRT is weaned as the renal function recovers — the falling creatinine, the rising urine output, the resolution of the indication. The CRRT-to-IHD transition is the typical pathway for the patient who is stabilising but not yet recovered: the CRRT is stopped, the IHD is started, and the modality is converted as the urine output returns. The premature cessation risks the rebound (the hyperkalaemia, the acidosis, the fluid overload); the prolonged CRRT in the recovering patient delays the mobilisation and the nutrition.[1][1]

The predictors of the successful CRRT cessation — the urine output over 400 mL/day (the strongest single predictor), the creatinine falling or stable, the absence of the oliguria, the improving metabolic control. The "crash and burn" (the recurrence of the uraemia or the hyperkalaemia within 48 h of the cessation) suggests the premature cessation.[1]

The workflow — starting, managing, switching

The trial cards — the evidence on the timing and the dose

[1] [1] [1] [1] [1] [1]

Clinical pearls

Additional red flags

References

1. [1]Kellum JA, Lameire N Acute kidney injury: an increasing global concern. Lancet, 2013.PMID 23727171
2. [2]Gaudry S, Hajage D, Schortgen F, et al Initiation Strategies for Renal-Replacement Therapy in the Intensive Care Unit. N Engl J Med, 2016.PMID 27181456
3. [3]The STARRT-AKI Investigators; Bagshaw SM, Wald R, et al Timing of Initiation of Renal-Replacement Therapy in Acute Kidney Injury. N Engl J Med, 2020.PMID 32668114
4. [4]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.PMID 27209269
5. [5]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.PMID 19846848
6. [6]Palevsky PM, Zhang JH, O'Connor TZ, et al (VA/NIH ATN) Intensity of renal support in critically ill patients with acute kidney injury. N Engl J Med, 2008.PMID 18492867
7. [7]Bellomo R, Kellum JA, Ronco C Acute kidney injury. Lancet, 2012.PMID 22617274
8. [8]Schortgen F, Bouadma L, Ladan S, et al (IDEAL-ICU) Timing of Renal-Replacement Therapy in Patients with Acute Kidney Injury and Sepsis. N Engl J Med, 2018.PMID 30304656
9. [9]Heintz BH, Matzke GR, Dager WE Antimicrobial dosing concepts and recommendations for critically ill adult patients receiving continuous renal replacement therapy or intermittent hemodialysis. Pharmacotherapy, 2009.PMID 19397464
10. [10]Schneider AG, Bellomo R, Bagshaw SM, et al Choice of renal replacement therapy modality and dialysis dependence after acute kidney injury: a systematic review and meta-analysis. Intensive Care Med, 2013.PMID 23443311