Critical Care Nephrology · Two-Week Intensive · Lecture 8 of 9

Kidney Replacement Therapy I

Access, modality and timing — three decisions made before a single litre of effluent is produced

40 minutes Nephrology Fellows Week 2

Based on Koyner, Handbook of Critical Care Nephrology (2021) · Ronco, Critical Care Nephrology 3e · NTUH Yunlin Branch

Learning objectives

By the end of this session you will be able to…

Choose site, length and calibre of acute access from the physics and the randomised data, and troubleshoot dysfunction and recirculation quantitatively.
Derive modality choice from transport mechanism — sieving, filtration fraction, dilution mode, membrane cut-off — rather than from local habit.
Quote what ELAIN, AKIKI, IDEAL-ICU, STARRT-AKI and AKIKI-2 each showed, and state precisely what remains unsettled.
Write a trigger list for watchful waiting, and a liberation plan that treats de-escalation as a prescription.
Frame an uncertain start as a time-limited trial with named milestones and a date.

Koyner Ch 29, 31–32, 56 · Ronco Ch 140, 144, 163, 165–167, 179, 183 · KDIGO AKI 2012 (KDIGO 2026 AKI/AKD guideline in public review)

Vascular access

Every dose you fail to deliver, every filter you lose and half the infections you cause were decided at the moment the catheter went in.

Koyner Ch 29 · Ronco Ch 167 · KDIGO 2012

“The access is the mother of dialysis.”

Not a central line: dual-lumen, 11–13.5 Fr, delivering 200–400 mL/min
Hagen–Poiseuille: resistance falls with the fourth power of radius — one French size beats any pump setting
Non-tunnelled is KDIGO's first choice — removed the day it is not needed
Tunnelled cuffed beyond 1–3 weeks: fewer infections
A fistula can run CRRT, but needle dislodgement in a proned patient is lethal

Blood-flow targets you must be able to feed

IHD 300–400 · SLED ≈200 · CRRT 150–250 mL/min.

Nothing matters if the blood pump cannot be fed.

Koyner Ch 29 · Koyner Ch 32 Table 32.1 (flow rates) · KDIGO AKI 2012 §5.4 · Al Rifai A et al. Hemodial Int. 2018;22:50 (AVF/AVG for CRRT)

Site selection: the order is evidence-based, not habit

Site	Rank	Mechanistic argument	Dysfunction
Right internal jugular	1st	Shortest, straightest path; compressible	6.6%
Femoral	2nd	Fastest in arrest; no pneumothorax risk	10.3%
Left internal jugular	3rd	Two near-right-angle turns — kinking, worse flow	19.5%
Subclavian	Last	Stenosis destroys future fistula territory	—

Parienti · Cathedia

n = 750, femoral vs jugular: no difference in bloodstream infection. Femoral was worse only in the highest BMI tertile, and better in the lowest.

Parienti JJ et al. JAMA. 2008;299:2413–2422 · Parienti JJ et al. Crit Care Med. 2010;38:1118–1125 (dysfunction, n=736) · KDIGO 2012 · Koyner Ch 29

Length, tip position and the physics of recirculation

Site	Length	Target tip	Recirculation
Right internal jugular	15–20 cm	Cavoatrial junction	≈10%
Left internal jugular	20–24 cm	Cavoatrial junction	≈10%
Femoral vein	≥24 cm	IVC above the confluence	≈8%
Lines reversed	—	—	20–30%

Why: recirculation is just-returned blood re-entering the arterial port. It rises when the tip lies in a low-flow segment.

Longer is better — and it was randomised

A longer catheter in a great thoracic vein prolonged filter survival by ~6.5 h.

delivered ≈ prescribed × ( 1 − R ) Femoral < 20 cm: ≈26% recirculation — shorter is worse.

Morgan D et al. Am J Kidney Dis. 2012;60:272–279 · Little MA, Conlon PJ, Walshe JJ. Am J Kidney Dis. 2000;36:1135–1139 · Ronco Ch 167 · Koyner Ch 29 Table 29.1

Putting it in, and keeping it alive

PlanSite, length, platelets/INR, pre-scan for thrombus.

→

BarrierMaximal barrier precautions.

→

Real-time ultrasoundNeedle tip visualised into the vein.

→

ConfirmBoth lumens aspirate and flush.

Every shift, every connection

Scrub the hub at every connect; never leave a hub open
Inspect the exit site
Chart blood flow, pressures, catheter-days
Ask every round: does this line still need to be in?

Locking solutions — what the evidence supports

Heparin vs saline: minimal difference — saline is a defensible default
4% citrate: comparable patency, less bleeding
Concentrated citrate: aspirate the lock, never flush it
No routine antibiotic locks

Rabindranath KS et al. Cochrane Database Syst Rev. 2011;(11):CD005279 · O'Grady NP et al. Clin Infect Dis. 2011;52:e162 · Lopez-Briz E et al. Cochrane 2018;7:CD008462 · Zhao Y et al. Am J Kidney Dis. 2014;63:479 · Koyner Ch 29

Complications — and the two errors you cannot take back

Mechanical · at insertion

Arterial cannulation, haematoma
Pneumothorax
Femoral: AV fistula
Guidewire arrhythmia

Infectious · over days

Exit-site infection, CRBSI
Risk accrues per catheter-day
Draw paired cultures before blaming the line
Never wire-exchange a suspected infected line

Thrombotic · over weeks

Mural and intraluminal thrombus
Host-vein thrombosis, central stenosis
More than half removed for a complication in one cohort

Two irreversible errors

1 · The convenient subclavian line — central venous stenosis 42% vs 10% against jugular access.
2 · Burning the arm before nephrology is called. No PICCs or cannulas in the non-dominant arm of anyone with CKD 4–5 — it is the future fistula.

Kairaitis LK, Gottlieb T. Nephrol Dial Transplant. 1999;14:1710 · Schillinger F et al. NDT. 1991;6:722 · Hernández D et al. J Am Soc Nephrol. 1998;9:1507 · Clark E et al. Can J Kidney Health Dis. 2016;3 · KDIGO 2012

Catheter dysfunction: a ladder, not a shrug

Define it numerically: arterial pressure <−250 or venous >+250 mmHg at target flow. Rising transmembrane pressure = a filter problem.

Position first — hip flexion, head turn, a kink. The commonest cause.

Volume — hypovolaemia collapses the vein around the arterial lumen.

Reverse the lines only as a bridge — it costs 20–30% recirculation.

Thrombolytic lock — an alteplase dwell for intraluminal thrombus.

Exchange or resite — a new site if infection is suspected.

Ronco Ch 167 · Koyner Ch 29 · Clark E et al. Can J Kidney Health Dis. 2016;3:2054358116669128

Modalities & the physics behind them

No modality has ever been shown in a randomised trial to save more lives than another — so choose the one whose transport physics matches the problem you are trying to solve.

Koyner Ch 32 · Ronco Ch 140, 150, 165–166, 176–179, 183

The alphabet, decoded once — and reported the ADQI way

IHD

3–4 h, blood 300–400, dialysate 500–800 mL/min. Pure diffusion.

SLED / PIRRT

6–12 h, blood ≈200. Diffusion, slowly.

SCUF

Plasma-water removal only — when volume is the entire problem.

CVVH

Convection alone. Best middle-molecule clearance.

CVVHD

Diffusion alone. Small-solute control, best filter life.

CVVHDF

Both: dialysate and replacement fluid.

No circuit, no anticoagulation, no disequilibrium.

Effluent

The ADQI dose unit — always state prescribed and delivered.

Neri M et al. Crit Care. 2016;20:318 (ADQI nomenclature & standardised reporting for KRT) · Koyner Ch 32 Table 32.1 · Ronco Ch 176–177

Four mechanisms, one membrane — why molecular size decides

Mechanism	Driving force	Clears	Falls off when
Diffusion	Concentration gradient; flux ∝ 1/√MW	Small solutes <500 Da	Above ~1 kDa
Convection	Transmembrane pressure dragging solvent	Small and middle molecules	The protein layer plugs the pores
Ultrafiltration	Hydrostatic pressure	Plasma water	Alone — no solute control
Adsorption	Binding to membrane and pores	Some hydrophobic peptides	Saturates within hours

SC = C_UF ÷ C_plasma; cut-off = MW at SC 0.1 Urea SC = 1.0. In plasma, adsorbed protein narrows the pores.

Ronco Ch 140 (cut-off, retention onset, high cut-off) · Ronco Ch 150, 165 · Brunet S et al. Am J Kidney Dis. 1999;34:486 · Koyner Ch 32

Pre- vs post-dilution: filter life bought with clearance

Filtration fraction = Q_UF ÷ [ Q_B × (1 − Hct) ] → keep ≤20–25% in post-dilution At FF 30% the haematocrit leaving the fibres is a third higher. Raising blood flow drops FF.

Pre-dilution

Replacement fluid enters before the filter
Costs roughly 15% of solute clearance for the same effluent volume
Lowers FF → longer filter life
Default in the clot-prone

Post-dilution

Replacement fluid enters after the filter
Most efficient per litre
Raises FF → shorter filter life
Unforgiving with under-delivery

Uchino S et al. Nephron Clin Pract. 2003;94:c94 · Ronco Ch 165 · Koyner Ch 32 Figure 32.1

CRRT vs IHD — and why "gentler" is the wrong word

Dimension	CRRT	IHD
Haemodynamic tolerance	Best — UF ≈2 mL/min	Worst — ≈12.5 mL/min
Net ultrafiltration control	Hourly	Bolus-like; balance drifts
Solute kinetics	Low hourly clearance	High clearance, peak–trough
Osmolality / ICP	Stable — preferred in brain injury	Rapid urea fall raises ICP
Anticoagulation	Citrate preferred	Often none
Rapid K⁺ or toxin removal	Slow	The correct tool
Mobility, theatre, imaging	Tethered	Free
Nutrition & blood-product volume	Unlimited	Constrained

The mechanism behind "haemodynamic tolerance"

Not gentle — balanced. UF stays below refill; urea falls slowly; cooling raises SVR. Much of the difference disappears with cool dialysate.

Koyner Ch 32 Tables 32.2, 32.4 · Ronco Ch 163, 166 (UF rate, osmolar and thermal balance) · KDIGO 2012 §5

Does modality change outcome? What is settled, what is not

Hemodiafe & CONVINT

Hemodiafe: CVVHDF vs protocolised IHD — no 60-day survival difference detected. CONVINT agrees. Both excluded the most shocked patients.

Observational · recovery signal

STARRT-AKI post hoc: less death or KRT dependence at 90 d (OR 0.81). AKIKI/IDEAL-ICU post hoc points the other way. Confounding runs both ways.

How to say this on rounds

CRRT is chosen for control — pressure, volume, osmolality — not for a survival curve. Until ICRAKI reports, defend the delivery.

Vinsonneau C et al. Lancet. 2006;368:379 · Schefold JC et al. Crit Care. 2014;18:R11 · Wald R et al. Intensive Care Med. 2023;49:1305 · Gaudry S et al. Crit Care. 2022;26:93 · Uchino S et al. JAMA. 2005;294:813 · Gaudry S et al. Crit Care Resusc. 2025 (ICRAKI protocol)

Choosing a modality: the one-slide algorithm

Fast removal needed (refractory K⁺, toxin) → IHD now, whatever the BP.

Vasopressor-dependent, brain-injured, on ECMO? → CRRT: osmolar stability, hourly UF titration.

Volume the only problem? → SCUF or diuretics first.

Stabilising, or the day needed for theatre? → SLED overnight.

Stable between sessions? → IHD.

No safe access? → acute PD if the abdomen allows.

Key point

Modality is a daily prescription. CRRT → SLED → IHD is the normal recovery trajectory.

KDIGO 2012 §5 · Lund A et al. Acta Anaesthesiol Scand. 2019;63:493 (ICP during IHD) · Vijayan A et al. Kidney Int Rep. 2018;3:456 · Koyner Ch 32

The other two: PIRRT and acute peritoneal dialysis

SLED / PIRRT — the pragmatic middle

Prescription: 6–12 h overnight; blood ≈200, dialysate ≈300 mL/min
Small-solute clearance without 4-h swings; frees the day
Honest gap: no consensus dose or frequency
In most units PIRRT substitutes for CRRT

Acute PD — misplaced, not obsolete

≈35 L/day in 2-L dwells; weekly Kt/V ≈3.5–4
Wins: no access, no anticoagulation, no water or electricity
Fails: unpredictable UF, hypercatabolism, open abdomen, raised IAP

Randomised comparisons

SLED vs CVVH: 90-day mortality NS. PD vs daily HD: equivalent mortality and recovery. All small, mostly single-centre.

Schwenger V et al. Crit Care. 2012;16:R140 · Zhang L et al. Am J Kidney Dis. 2015;66:322 · Chionh CY et al. Clin J Am Soc Nephrol. 2013;8:1649 · Cullis B et al. Perit Dial Int. 2014;34:494 (ISPD) · Ronco Ch 179, 183 · Koyner Ch 32

Timing: starting, and stopping

Five randomised trials asked whether starting earlier helps. The answer is no — but the wait has a floor, and the exit has been studied even less than the entrance.

Koyner Ch 31, 51 · Ronco Ch 144, 163–164

The easy decision, and the hard one

	Urgent	Working threshold
A	Acidaemia	pH ≤7.20 or HCO₃⁻ ≤12, refractory
E	Electrolytes	K⁺ ≥6.0 refractory, or any ECG change
I	Intoxication	A dialysable toxin (EXTRIP)
O	Overload	Pulmonary oedema refractory to diuretics
U	Uraemia	Encephalopathy, pericarditis or bleeding

…but AEIOU is rarely why KRT starts

Emergency indications are not the commonest trigger — oligoanuria and balance trends are.

Ask instead: is demand exceeding this kidney's capacity?

KDIGO 2012 · STARRT-AKI. N Engl J Med. 2020;383:240 · Gaudry S et al. N Engl J Med. 2016;375:122 · Mehta RL. Blood Purif. 2001;19:227 · Bagshaw SM et al. Clin J Am Soc Nephrol. 2019;14:496 · Koyner Ch 31 · Ronco Ch 163

The trials, part 1 — the one positive result, and the answer to it

ELAIN · JAMA 2016

Single centre, mostly surgical. CRRT within 8 h vs at stage 3: 90-day mortality 39.3% vs 54.7%. Separation was only 21 h, and it has never been replicated.

AKIKI · NEJM 2016

Stage 3 on ventilation or vasopressors. KRT within 6 h vs deferral: 60-day mortality 48.5% vs 49.7%. 49% of the delayed arm never received KRT.

Zarbock A et al. JAMA. 2016;315:2190–2199 · Gaudry S et al. N Engl J Med. 2016;375:122–133 · Koyner Ch 31 Table 31.1

The trials, part 2 — the two that settled the question

IDEAL-ICU · NEJM 2018

Septic shock. KRT within 12 h vs after 48 h: stopped for futility; 90-day mortality 58% vs 54%. The early-support hypothesis fails.

STARRT-AKI · NEJM 2020

168 centres, stage 2–3 AKI, enrolled only where equipoise existed. 90-day mortality 43.9% vs 43.7%. Among survivors, KRT dependence 10.4% vs 6.0%.

Barbar SD et al. N Engl J Med. 2018;379:1431–1442 · STARRT-AKI Investigators. N Engl J Med. 2020;383:240–251 · Koyner Ch 31

The floor, the pooled estimate, and one loose end

AKIKI-2 · Lancet 2021

Patients already at the AKIKI delayed threshold, randomised to start now or wait. No gain in KRT-free days; mortality rose with more delay (aHR 1.65). Deferral cannot be extended indefinitely.

IPD meta-analysis · Lancet 2020

9 trials, n = 1879: 28-day mortality 44% delayed vs 43% early — precise, and null.

What STARRT-AKI did not settle

A 2025 post hoc found the dependence signal attenuated and non-significant. Still unsettled: no-equipoise patients, stage 1 AKI, cardiac surgery.

Gaudry S et al. Lancet. 2021;397:1293–1300 · Gaudry S et al. Lancet. 2020;395:1506–1515 · McCoy IE et al. Clin J Am Soc Nephrol. 2025;20:601–607 (STARRT-AKI post hoc; doi:10.2215/CJN.0000000672) · Koyner Ch 31

The synthesis: watchful waiting is an active prescription

Write this in the notes	Threshold
Potassium	≥6.0, or rising despite therapy
Acid–base	pH ≤7.20 or HCO₃⁻ ≤12
Oxygenation	PaO₂/FiO₂ ≤200, or refractory oedema
Cumulative balance	>10% of body weight
Uraemia	Encephalopathy, pericarditis or bleeding
Duration	Oliguria beyond 72 h with no recovery

What makes it "watchful"

A named threshold list in the notes
A daily documented decision — including "not today"
An access plan ready, so starting takes 30 minutes
Diuretics, nutrition and ventilation optimised while you wait

STARRT-AKI. N Engl J Med. 2020;383:240 · Gaudry S et al. Lancet. 2021;397:1293 · KDIGO AKI 2012 · KDIGO 2026 AKI/AKD guideline public review draft, March 2026 (kdigo.org) · Koyner Ch 31

2026 KDIGO draft: defer by default, but define the floor

Urgent: start now

Refractory hyperkalaemia, severe acidaemia, dialysable intoxication
Diuretic-resistant pulmonary oedema
Uraemic bleeding, pericarditis or encephalopathy

Relative: prepare and reassess

Persistent stage 3 AKI or prolonged oligoanuria
Other organ dysfunction worsened by AKI
Demand–capacity mismatch

Default when neither applies

Deferred strategy with a named trigger list
Place access early when an urgent start is plausible
Don't convert a number into an automatic start

How to teach this slide

The draft makes "watch and wait" an active prescription, and keeps the decision patient-specific.

KDIGO. 2026 Clinical Practice Guideline for AKI and AKD, public-review draft, March 28, 2026: Table 41 and §§5.2–5.4 (draft; KDIGO 2012 remains operative; wording may change before final publication)

Sharpening the decision: the furosemide stress test and biomarkers

Furosemide stress test

Furosemide 1.0 mg/kg (naive) or 1.5 mg/kg (prior exposure), euvolaemic, replacing losses for 6 h
UO <200 mL/2 h predicts progression (AUC ≈0.87)
A flat response means secure access — it is not an order to start

Biomarkers

Cell-cycle arrest/damage markers stratify risk (Lecture 4); evidence insufficient to guide KRT initiation
ELAIN used NGAL >150 only as an enrichment criterion
No biomarker has improved outcomes

Key point

These tools move the probability of recovery. They don't answer whether acting earlier helps — it did not.

Chawla LS et al. Crit Care. 2013;17:R207 · Koyner JL et al. J Am Soc Nephrol. 2015;26:2023 · Klein SJ et al. Intensive Care Med. 2018;44:323 · Koyner Ch 16, 31

Two ways waiting goes wrong — and one way starting does

Pitfall · waiting while the balance climbs, then removing it too fast

Fluid accumulation at initiation and cessation tracks mortality — overload at cessation ≈ 2.5× odds (PICARD). But correction has a ceiling: net UF >1.75 mL/kg/h vs <1.01 carried HR 1.51 for death (RENAL); every extra 0.5 mL/kg/h added ~8–13% hazard. At 80 kg, >1.75 mL/kg/h means >140 mL/h. Plan fluid removal over days, not shifts.

Pitfall · calling it waiting when nobody is watching

Watchful waiting without a written trigger list, daily review and a ready access plan is not the strategy the trials tested — it is drift.

Bouchard J et al. Kidney Int. 2009;76:422–427 (PICARD) · Murugan R et al. JAMA Netw Open. 2019;2:e195418 (RENAL secondary) · Ronco Ch 163 · Koyner Ch 31

Stopping: predicting recovery, and the case for doing less

Urine output off diuretics is the strongest simple predictor; cut-off around 400–500 mL/day
Diuretics degrade the signal — a litre on furosemide is not recovered clearance
Timed creatinine clearance is objective: ATN stopped support above 20 mL/min
A pre-treatment creatinine that falls between sessions signals returning function

Why late stopping costs you

Every extra day buys catheter risk, anticoagulation and immobility. Each hypotensive episode is a fresh ischaemic hit to a repairing kidney.

LIBERATE-D · JAMA 2026

The first RCT of de-escalation: recovery at discharge 64% vs 50%. Unblinded, and the adjusted analysis was non-significant.

Uchino S et al. Crit Care Med. 2009;37:2576 · Palevsky PM et al. N Engl J Med. 2008;359:7 (ATN) · Liu KD et al. JAMA. 2026;335:326–335 (published online 2025; doi:10.1001/jama.2025.21530) · Ronco Ch 144 · Koyner Ch 31, 51

How to wean, in the order you should do it

Stop taking fluidNet UF to zero at target volume; see whether balance holds on diuretics.

→

Step down intensityCRRT → SLED → alternate-day IHD, guided by chemistry.

→

Space the treatmentsExtend the interval as urine output allows.

→

Hold and observeStop in hospital and watch for 48–72 h.

Practical rules

Keep the catheter until the hold has succeeded, then remove it the same day. About 6–10% of survivors stay dialysis-dependent at 90 days — every discharge needs written follow-up.

Ronco Ch 144 · STARRT-AKI. N Engl J Med. 2020;383:240 · Uchino S et al. Crit Care Med. 2009;37:2576 · Koyner Ch 51 · NTUH Yunlin unit practice

Shared decisions & time-limited trials

"Can we dialyse this patient?" is a technical question with an easy answer; "should we?" is the one that needs a structure.

Koyner Ch 56 · Ronco Ch 148 · RPA guideline 2010

Clinical momentum, and the numbers that should slow it

Dialysis decisions arrive bundled with intubation and tracheostomy. Clinical momentum accumulates interventions without revisiting the direction
Older patients starting dialysis in hospital do badly: survival 55.8% at 6 months, 45.5% at 1 year
AKI on CKD frequently does not recover
Resist the technological imperative: being able to dialyse is not a reason to dialyse

Ask before you consent

Baseline kidney function · comorbidity burden · functional status · advance directives · quality vs length of life

Kruser JM et al. Ann Am Thorac Soc. 2017;14:426 · Wachterman MW et al. JAMA Intern Med. 2019;179:987 · Bagshaw SM et al. Clin J Am Soc Nephrol. 2019;14:496 · STARRT-AKI. N Engl J Med. 2020;383:240 · Koyner Ch 56 Box 56.1

The time-limited trial: four steps you can run this week

PrepareAgree the prognosis, what KRT can achieve, and the milestones.

→

CommunicateAsk–Tell–Ask. Elicit values, propose a time-limited trial.

→

ConductDeliver full therapy; review daily.

→

ConcludeMeet on the agreed date. Met → continue; not met → comfort care.

Opening

Use dialysis for five days; meet Friday.

Eliciting values

What would he say if he could see himself here?

Naming the worry

I worry we are keeping his kidneys going in a body that is not recovering.

Setting the frame

A trial succeeds either way.

Scherer JS, Holley JL. Clin J Am Soc Nephrol. 2016;11:344 · Renal Physicians Association. Shared Decision-Making in the Appropriate Initiation of and Withdrawal from Dialysis. 2nd ed. 2010 · Koyner Ch 56

Withholding, withdrawing and conservative kidney management

When it is ethically appropriate to withhold or withdraw

A patient with capacity who, fully informed, refuses or asks to stop
No capacity, previously refused in an advance directive
No capacity, proxy declines
Irreversible profound neurological impairment
Withholding and withdrawing are ethically equivalent

Conservative and palliative kidney management

Active care without dialysis: volume and symptom control, potassium management, honest prognostication
Anticipate uraemia and hyperkalaemia after withdrawal and prescribe for dyspnoea, agitation and pain in advance
Involve palliative care early

Renal Physicians Association guideline (2nd ed, 2010) Table 56.2 · Chong K et al. Clin J Am Soc Nephrol. 2017;12:1744 · Koyner Ch 56 · Ronco Ch 148

Cases & wrap-up

Three referrals, three different questions: when, which, and whether.

Koyner Ch 29, 31–32, 56

Case 1 · Septic AKI, stage 3, no emergency

66 M, 78 kg, pneumonia, ICU day 3. Noradrenaline weaning, lactate normal. Loop diuretic — little effect.

Cr 1.1 → 3.6UO 12 mL/hK 5.2pH 7.31HCO₃ 18P/F 240+3.4 LNE 0.06

Start KRT today, or not?

Answer — watchful waiting

Do not start. Same mortality as STARRT-AKI standard arm; 10.4% vs 6.0% dependence favours waiting. 49% of AKIKI delayed never needed KRT.

What makes it active

Written triggers · optimise · FST if unclear · daily review with ready access. AKIKI-2: escalate at 72 h.

STARRT-AKI. N Engl J Med. 2020;383:240 · Gaudry S et al. N Engl J Med. 2016;375:122 (AKIKI) · Barbar SD et al. N Engl J Med. 2018;379:1431 (IDEAL-ICU) · KDIGO 2012

Case 2 · Severe traumatic brain injury with AKI and overload

27 M, 70 kg, day 4 after a motorcycle collision. Craniectomy, ICP monitored, on hypertonic saline. Rhabdomyolysis-associated AKI, now anuric and ventilated. The surgical team wants 4 L removed "as soon as possible".

ICP 21 mmHgNa 152 mmol/LCr 4.4 mg/dLK 5.8Balance +6.2 LP/F 180CK 42 000 U/L

Which modality, which two numbers must you protect, and how fast may you remove the 4 L?

Think 45 seconds — modality, two targets, one rate

Lund A et al. Acta Anaesthesiol Scand. 2019;63:493 (ICP during IHD) · Murugan R et al. JAMA Netw Open. 2019;2:e195418 (RENAL secondary) · Koyner Ch 32 · Ronco Ch 166

Case 3 · The referral where the question is "whether"

84 M, CKD stage 4, moderate dementia, dependent for washing and dressing. Admitted with aspiration pneumonia and shock; ventilated, on two vasopressors, oliguric 30 h. No advance directive. The family asks you to "do everything".

Cr 2.8 → 5.1UO 8 mL/hK 5.4pH 7.26Balance +5 LKarnofsky ~40%

What do you offer, and how do you say it?

Scherer JS, Holley JL. Clin J Am Soc Nephrol. 2016;11:344 · Renal Physicians Association guideline (2nd ed, 2010) · Koyner Ch 56

His baseline creatinine is 2.8 mg/dL and he has had three admissions this year for fluid overload. Model answer: offer a time-limited trial rather than a binary yes/no. Prepare — agree with the ICU team what KRT can plausibly achieve and by when, e.g. vasopressor independence and improving oxygenation within 5–7 days. Communicate with Ask–Tell–Ask and state the prognosis honestly: he carries several independent predictors of 1-year mortality (advanced age, ADL dependence, inpatient initiation, multiple comorbidities), and AKI on CKD 4 makes permanent dialysis dependence likely. Set the review date before starting, document it, involve palliative care in parallel rather than after failure. Teaching point: "do everything" almost always means "do not abandon us" — a trial with a date answers that without committing him to indefinite organ support. You may ethically decline to offer dialysis when burdens clearly outweigh benefits, but that conversation lands better after a trial has been offered and reviewed. 2 min.

Quick poll

You are asked to justify a modality choice. Which statement is best supported by randomised evidence?

A. CRRT improves 90-day survival over IHD in vasopressor-dependent AKI · B. Convective therapy (CVVH) clears middle molecules better than CVVHD and this translates into lower mortality in sepsis · C. Protocolised IHD with cool dialysate and slow ultrafiltration achieves survival no different from CVVHDF in AKI with multiorgan failure · D. Pre-dilution CVVH delivers more solute clearance per litre of effluent than post-dilution

Hands up — then name the trial that decides it

Vinsonneau C et al. Lancet. 2006;368:379 (Hemodiafe) · Schefold JC et al. Crit Care. 2014;18:R11 · Wald R et al. Intensive Care Med. 2023;49:1305 · Koyner Ch 32

Answer C — Hemodiafe (Lancet 2006, n=360) randomised AKI with multiorgan failure to CVVHDF vs protocolised IHD and found no 60-day survival difference; CONVINT agrees. A is the commonest specialist error: the CRRT recovery/survival signal is observational and points in both directions (STARRT-AKI secondary analysis favours CRRT for 90-day KRT dependence, OR 0.61; the AKIKI/IDEAL-ICU secondary analysis favours IHD for 60-day survival, HR 1.26) — and ICRAKI is being run precisely because the question is open. B is half true and half false: convection does clear middle molecules better (higher sieving coefficients across the 500–5000 Da range), but OMAKI and a 19-trial meta-analysis show no mortality benefit; mechanism is not outcome. D inverts the physics: pre-dilution costs about 15% of solute clearance per litre of effluent because the blood is diluted before it is cleaned — it buys filter life, not clearance. Teaching point: every one of these distractors is something you will hear said confidently on a ward round. 2 min.

Key takeaways

Access is Poiseuille: right IJ first, femoral second (≥24 cm), left IJ third, subclavian never.
Diffusion clears <500 Da; convection adds middle molecules. Pre-dilution trades ~15% for filter life.
CRRT is tolerated via volume, osmolar and thermal balance; with the same protections IHD performs no worse.
Early initiation does not save lives, and 42% of delayed patients are never dialysed — but the wait has a floor.
Watchful waiting = trigger list, daily review, ready access.
Stopping is a prescription too. Frame uncertain starts as time-limited trials.

References & further reading

Where to go deeper

Parienti JJ, et al. Femoral vs jugular. JAMA. 2008;299:2413.
Morgan D, et al. Catheter length RCT. Am J Kidney Dis. 2012;60:272.
Neri M, et al. ADQI nomenclature. Crit Care. 2016;20:318.
Vinsonneau C, et al. Hemodiafe. Lancet. 2006;368:379.
Zarbock A, et al. ELAIN. JAMA. 2016;315:2190.
Gaudry S, et al. AKIKI. NEJM. 2016;375:122.
STARRT-AKI Investigators. NEJM. 2020;383:240 · AKIKI-2. Lancet. 2021;397:1293.
Murugan R, et al. Net UF rate. JAMA Netw Open. 2019;2:e195418.

Full list: (1) Parienti JAMA 2008;299:2413–2422 + Crit Care Med 2010;38:1118–1125; (2) Morgan AJKD 2012;60:272–279; (3) Neri ADQI nomenclature (Crit Care 2016;20:318); (4) Hemodiafe (Lancet 2006;368:379–385); (5) Wald STARRT-AKI CRRT-vs-IHD (Intensive Care Med 2023;49:1305–1316), Gaudry Crit Care 2022;26:93; (6) ELAIN (JAMA 2016;315:2190–2199); (7) AKIKI (NEJM 2016;375:122–133), IDEAL-ICU (NEJM 2018;379:1431–1442); (8) STARRT-AKI (NEJM 2020;383:240–251); (9) Gaudry IPD meta-analysis (Lancet 2020;395:1506–1515), AKIKI-2 (Lancet 2021;397:1293–1300), McCoy (CJASN 2025;20:601–607); (10) RENAL net UF secondary (JAMA Netw Open 2019;2:e195418); (11) LIBERATE-D (JAMA 2026;335:326–335), Schwenger SLED vs CVVH (Crit Care 2012;16:R140), Chionh PD for AKI (CJASN 2013;8:1649–1660); (12) Scherer & Holley time-limited trials (CJASN 2016;11:344–353), RPA shared-decision guideline 2e 2010; (13) KDIGO 2012 (Kidney Int Suppl;2:1–138), KDIGO 2026 AKI/AKD draft; (14) Koyner Handbook Ch 16, 29, 31–32, 51, 56; Ronco Critical Care Nephrology 3e Ch 140, 144, 148, 150, 163, 165–167, 179, 183.

Critical Care Nephrology · Two-Week Intensive

Thank you

Questions & discussion — tomorrow we prescribe the therapy we have just decided to give.

Next: Lecture 09 — KRT II: Dose, Anticoagulation, Complications & Recovery