Critical Care Nephrology · Two-Week Intensive · Lecture 3 of 9

Fluids, Vasopressors, Diuretics & Blood Products

Prevention and conservative management of AKI — prescribing the four things we give every ICU patient without thinking

40 minutes Nephrology Fellows Week 1

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

Learning objectives

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

Choose a resuscitation fluid from its mechanism and its trial data — and name the four times saline still wins.
Prescribe a volume, an endpoint and a stop rule — then run evacuation as deliberately.
Run a furosemide stress test, and work the diuretic-resistance ladder to its end.
Write vasoactive orders in norepinephrine-equivalent doses, and treat HRS-AKI, cardiorenal syndrome and intra-abdominal hypertension by mechanism.
Apply restrictive transfusion with its exceptions, manage uraemic bleeding, and feed the AKI patient.

Koyner Ch 8–15, 40, 42–43, 46–47 · KDIGO AKI 2012 · Surviving Sepsis Campaign 2026 · WSACS 2013

Fluids: which one

Every fluid is a drug with a composition, a dose, a distribution volume and a toxicity — start prescribing it that way.

Koyner Ch 10 · Ronco Ch 18

Revised Starling, the glycocalyx, and what is actually in the bag

~25%of a crystalloid bolus stays intravascular

1 : 1.4albumin:crystalloid expansion ratio

~2 µmglycocalyx — the real oncotic barrier

Once the glycocalyx is shed, the colloid advantage disappears

Fluid	Na⁺	Cl⁻	Note
Plasma	140	100–104	The target
0.9% saline	154	154	~50 mmol/L excess Cl⁻
Lactated Ringer's	130	109	273 mOsm/L — hypotonic in vivo
Plasma-Lyte 148	140	98	Acetate/gluconate buffered; no Ca²⁺
Albumin 4–5%	~140	~128	SAFE fluid; a slow, expensive crystalloid

Woodcock TE, Woodcock TM. Br J Anaesth. 2012;108:384–394 · Levick JR, Michel CC. Cardiovasc Res. 2010;87:198 · Koyner Ch 10 (Table 10.2) · Ronco Ch 18 · values in mmol/L, osmolarity mOsm/L

Chloride: the mechanism behind the trials

Saline delivers ~50 mmol more chloride per litre than plasma
Distal chloride → macula densa → tubuloglomerular feedback → afferent vasoconstriction → GFR falls
In isolated kidneys, chloride reduced GFR; equimolar sodium did not
Post-operative Cl⁻ >110: longer stay, higher mortality

SID saline = 154 − 154 = 0 · SID plasma ≈ 40 → hyperchloraemic normal-anion-gap acidosis
Day-2 falling bicarbonate is usually the fluid — check the chloride.

The before-and-after study that started it

Yunos: chloride-liberal then chloride-restrictive eras. Injury/failure AKI fell 14% → 8.4%; adjusted OR for KRT 0.52. Sequential, unblinded.

Yunos NM et al. JAMA. 2012;308:1566–1572 · Wilcox CS. J Clin Invest. 1983;71:726–735 · McCluskey SA et al. Anesth Analg. 2013;117:412 · Koyner Ch 10

Balanced vs saline: four trials, one meta-analysis, one honest answer

Trial	n	Key result
SMART 2018	15,802	MAKE30 14.3% vs 15.4% (OR 0.90)
SALT-ED 2018	13,347	MAKE30 4.7% vs 5.6% (aOR 0.82)
BaSICS 2021	11,052	90-day mortality 26.4% vs 27.2% — neutral; TBI favoured saline
PLUS 2022	5,037	Mortality 21.8% vs 22.0% — neutral
BEST-Living IPD 2024	34,450	90-day mortality RR 0.96; benefit probability 89.5%

Read the composite honestly

MAKE30 = death, new KRT, or Cr >200%; AKI components alone NS
"Balanced prevents AKI" oversells

What to say on rounds

A small absolute benefit with low probability of harm. Balanced by default, saline by indication.

Semler MW et al. N Engl J Med. 2018;378:829–839 · Self WH et al. N Engl J Med. 2018;378:819–828 · Zampieri FG et al. JAMA. 2021;326:818–829 · Finfer S et al. N Engl J Med. 2022;386:815–826 · Hammond NE et al. Lancet Respir Med. 2024;12:29–38 (BEST-Living) · Surviving Sepsis Campaign 2026, Crit Care Med. 2026;54:725–812 · Koyner Ch 8, 10

Four times 0.9% saline is still the right prescription

Hypochloraemic alkalosis

After vomiting, NG loss or aggressive diuresis — the chloride is the therapy.

Brain injury & raised ICP

Ringer's (273 mOsm/L) is functionally hypotonic and worsens cerebral oedema. Use saline.

Hypercalcaemia & blood lines

Ringer's calcium worsens hypercalcaemia and chelates citrate in a blood line.

Significant liver failure

Lactate needs hepatic conversion — choose acetate-buffered (Plasma-Lyte) or saline; Ringer's lactate ≠ Plasma-Lyte.

Two myths worth killing

Ringer's does not cause lactic acidosis — it is sodium lactate. Nor is it contraindicated in hyperkalaemia.

Koyner Ch 10 · Zitek T et al. J Emerg Med. 2018;55:313 · O'Malley CMN et al. Anesth Analg. 2005;100:1518 · Zampieri FG et al. JAMA. 2021

Albumin: three trials that said no, and three indications that say yes

SAFE · ALBIOS · ATTIRE

SAFE (n = 6,997): mortality 20.9% vs 21.1%.

ALBIOS: no difference. ATTIRE: cirrhosis composite 29.7% vs 30.2%.

Where albumin does harm

TBI — contraindicated: SAFE mortality 33.2% vs 20.4%
Hyperoncotic 20–25% in shock: more Cr doubling
Serum-target albumin failed twice — give it for an indication, never a number

SBP — 1.5 g/kg day 1 + 1 g/kg day 3.

Paracentesis >5 L — 6–8 g/L.

HRS-AKI — 1 g/kg/day × 2 d to diagnose, then 20–40 g/day with a vasoconstrictor.

Finfer S et al. N Engl J Med. 2004;350:2247 · SAFE-TBI. N Engl J Med. 2007;357:874 · Caironi P et al. N Engl J Med. 2014;370:1412 · China L et al. N Engl J Med. 2021;384:808–817 (ATTIRE) · Sort P et al. N Engl J Med. 1999;341:403 · Caraceni P et al. Lancet. 2018;391:2417 · Schortgen F et al. Intensive Care Med. 2008;34:2157 · Surviving Sepsis Campaign 2026, Crit Care Med. 2026;54:725–812 · Koyner Ch 10, 42

Numbers behind the boxes: ALBIOS n = 1,818 (no mortality or AKI difference); ATTIRE n = 777 with more pulmonary oedema; SAFE-TBI RR 1.63; hyperoncotic albumin in shock OR 5.99 for creatinine doubling or dialysis; SSC 2026 says crystalloid alone. Albumin caps at 100 g/day for the HRS diagnostic challenge; SBP albumin cut renal impairment 33% → 10%; outpatient albumin (ANSWER) reduced SBP, HRS and kidney dysfunction. 2 min. ALBIOS raised MAP and lowered net fluid balance without touching mortality — the physiology of hyperoncotic albumin in sepsis is real, the outcome signal is not. ATTIRE's harm was mainly pulmonary oedema — hence the respiratory-reserve caveat on every HRS rung. SBP: mortality 29% → 10%, the highest-yield albumin order. HRS diagnosis: no response + bland sediment + no proteinuria = diagnosis by exclusion. Timing cue: 2 min.

Starches and gelatins: settled, and settled against

6S · N Engl J Med 2012

804 severe sepsis: 90-day mortality 51% vs 43%; more KRT.

CHEST · N Engl J Med 2012

7,000 ICU patients: mortality NS, more KRT 7.0% vs 5.8%.

CRISTAL · JAMA 2013

Colloid vs crystalloid: no 28-day difference; the 90-day signal was an unblinded secondary endpoint.

Pitfall — the "shortcut" colloid

HES causes osmotic nephrosis plus coagulopathy. KDIGO recommends against synthetic colloids. Gelatins are not the safe compromise marketed.

Perner A et al. N Engl J Med. 2012;367:124 · Myburgh JA et al. N Engl J Med. 2012;367:1901 · Annane D et al. JAMA. 2013;310:1809 · KDIGO AKI 2012 · Koyner Ch 8, 10

How much — and preventing the injury

Restrictive and liberal strategies barely differ in mortality; what kills is the fluid nobody ever decided to give.

Koyner Ch 8–10, 43, 46–47 · Ronco Ch 18

De-resuscitation: the phase we skip

Almost all ICU fluid is given in Stabilisation and Evacuation — and almost nobody prescribes the E phase.

CountYesterday's true intake. Units under-count by 1–2 L/day.

→

SubtractConcentrate infusions, go enteral, stop maintenance.

→

Prescribe a balance"−1 L" is an order, "even" is not. Negative by default once shock resolves.

→

Choose the toolLoop diuretic first; ultrafiltration only when the nephron cannot cope.

Key point

Fluid is prescribed in mL and removed in L — give evacuation the same daily target as the vasopressor.

Malbrain MLNG et al. Ann Intensive Care. 2018;8:66 · Liu KD et al. Crit Care Med. 2011;39:2665 · Koyner Ch 10 (Table 10.5) · Ronco Ch 18

How much: the trials are flat, the accumulation data are not

Trial	Volume separation	Outcome
CLASSIC 2022	1,798 vs 3,811 mL	90-day mortality 42.3% vs 42.1%
CLOVERS 2023	1,267 vs 3,400 mL	90-day death 14.0% vs 14.9%

The honest reading

Within the range tested, the exact volume does not decide survival.

>10%accumulation — the mortality threshold

~2×odds of death with overload at KRT start

−136 vs +6,992mL net balance, FACTT

Pitfall — chasing urine output with boluses in established ATN

A bolus "for the urine output" produces neither
Not responsive and not hypovolaemic? The fluid reflex is the diagnosis

Meyhoff TS et al. N Engl J Med. 2022;386:2459–2470 · Shapiro NI et al. N Engl J Med. 2023;388:499–510 · Bouchard J et al. Kidney Int. 2009;76:422–427 (PICARD) · ARDSNet FACTT. N Engl J Med. 2006;354:2564 · Teixeira C et al. Crit Care. 2013;17:R14 · Surviving Sepsis Campaign 2026, Crit Care Med. 2026;54:725–812 · Koyner Ch 8–10

Assessing responsiveness: the numbers, and where each test fails

Fluid challenge: 4 mL/kg over 10–15 min · positive = ≥10% SV rise
No pre/post SV measurement = fluid, not a challenge.

Test	Threshold	Where it fails
Passive leg raise (ΔSV)	≥10%; sens ~85%, spec ~91%	Direct CO needed; invalid with IAH
PPV / SVV	>12–13%; AUC 0.94 / 0.84	Sinus rhythm, TV ≥8; IAH inflates PPV
End-expiratory occlusion	15 s hold; ΔCI ≥5%	Needs a tolerated hold
Central venous pressure	AUC ≈ 0.56	Not for responsiveness; marks congestion

Testing changes what you give

FRESH: PLR-guided resuscitation cut 72-h balance ~1.4 L and KRT.

Monnet X, Marik PE, Teboul JL. Ann Intensive Care. 2016;6:111 · Marik PE et al. Crit Care Med. 2013;41:1774 · Douglas IS et al. Chest. 2020;158:1431–1445 (FRESH) · Koyner Ch 1, 10

The KDIGO prevention bundle — staged, not one-size

High risk, no AKI yet — stop nephrotoxins; secure volume and perfusion pressure; monitor Cr and UO; avoid hyperglycaemia.

Stage 1 — plus a non-invasive workup: sediment, ultrasound, drug review, bladder pressure.

Stage 2 — dose against the current, non-steady-state GFR; avoid nephrotoxic imaging.

Stage 3 — avoid subclavian lines if future access matters.

The risk factors worth memorising

Age, CKD, diabetes, heart failure · sepsis, high severity score, vasopressors, nephrotoxins, emergency surgery, long bypass time.

KDIGO AKI 2012 · Cartin-Ceba R et al. Crit Care Res Pract. 2012;2012:691013 · Koyner Ch 8

Does the bundle work — and what pressure? Four trials

Trial	Design	Key result
PrevAKI 2017	276 cardiac surgery, biomarker-triggered bundle	AKI at 72 h 55% vs 72%; stage 2/3: 30% vs 45%
BigpAK 2018	121 major abdominal surgery	Stage 2/3: 7% vs 20%
INPRESS 2017	298 high-risk surgery	SBP within 10% of baseline: kidney dysfunction 32.7% vs 49.0%
SEPSISPAM 2014	776 septic shock, MAP 80–85 vs 65–70	Mortality NS; chronic-HTN stratum: less KRT

Meersch M et al. Intensive Care Med. 2017;43:1551–1561 · Göcze I et al. Ann Surg. 2018;267:1013 · Futier E et al. JAMA. 2017;318:1346 · Asfar P et al. N Engl J Med. 2014;370:1583–1593 · Koyner Ch 8

What does not prevent AKI — and why we keep trying

Intervention	Theory	Verdict
"Renal-dose" dopamine	D₁ vasodilation, natriuresis	61 RCTs: more urine, no change in KRT or death. Dead.
Fenoldopam	Selective D₁ agonist	Stopped for futility; more hypotension.
Furosemide	Less mTAL O₂ demand	No prevention — volume control only.
Mannitol	Tubular flushing	No prevention outside raised ICP; hyperosmolar AKI.
N-acetylcysteine + bicarbonate	Antioxidant; alkalinised urine	PRESERVE and AMACING: stop ordering both.
Statins	Endothelial stabilisation	No ICU benefit demonstrated.
Levosimendan, EPO, selenium, RIP	Various	All negative in powered trials.

Friedrich JO et al. Ann Intern Med. 2005;142:510 · De Backer D et al. N Engl J Med. 2010;362:779 (SOAP II) · Bove T et al. JAMA. 2014;312:2244 · Weisbord SD et al. N Engl J Med. 2018;378:603 (PRESERVE) · Garg AX et al. JAMA. 2014;312:2254 (POISE-2) · Koyner Ch 8 (Table 8.3)

Three populations where the fluid rules change

Burns

Rule of 10s: mL/h = %TBSA × 10; titrate to UO
Ivy index: >250 mL/kg in 24 h = runaway resuscitation
Vitamin C risks oxalate nephropathy
CKRT 20–30 mL/kg/h

Trauma

AKI in ~20–24% of critically ill trauma patients
Transfusion volume is the main modifiable risk factor
Damage-control resuscitation did not increase AKI
Crush injury: relieve the compartment first

Intra-abdominal hypertension

IAH >12; ACS >20 with organ failure — AKI comes first
Measure intravesically, supine, end-expiration
APP = MAP − IAP — target ≥60
IAH independently predicts ICU mortality

Koyner Ch 46 (burns) · Ch 47 (trauma) · Ch 43 (ACS) · Kirkpatrick AW et al. Intensive Care Med. 2013;39:1190–1206 (WSACS) · Cheatham ML et al. J Trauma. 2000;49:621 · Malbrain MLNG et al. Minerva Anestesiol. 2014;80:293

Drugs for the failing circulation

Diuretics buy volume control and prognostic information — never renal protection. Vasoactives buy a perfusion pressure — never a kidney.

Koyner Ch 9, 12, 13, 40, 42

Loop diuretics: the pharmacology, and the two honest columns

Loops are >95% albumin-bound — OAT1/OAT3 secretion delivers them to NKCC2
Uraemic anions and hypoalbuminaemia impair OAT secretion
Threshold and ceiling — double the dose, not the interval
Braking: DCT hypertrophies — hence the thiazide
Bioavailability: furosemide 40–60%, torsemide >90%

Doses that matter

Furosemide 40 mg IV ≈ bumetanide 1 mg ≈ torsemide 20 mg. Weight-based: 1 mg/kg naive, 1.5 mg/kg after prior exposure.

What they deliver

Volume control, lower CVP, better RPP.

What they do not deliver

No prevention, no treatment of established AKI.

How to reconcile the two columns

KDIGO: against diuretics to prevent or treat AKI, for volume management. Prescribe to volume, never to creatinine.

Ellison DH. Clin J Am Soc Nephrol. 2019;14:1248 · Hoorn EJ, Ellison DH. Am J Kidney Dis. 2017;69:136 · Ellison DH et al. J Clin Invest. 1989;83:113 · KDIGO AKI 2012 · Mehta RL et al. JAMA. 2002;288:2547 · Teixeira C et al. Crit Care. 2013;17:R14 · Koyner Ch 9, 12 (Table 12.1)

Threshold and ceiling in full: below the threshold there is no natriuresis at all, and above the ceiling there is no extra — hence doubling the dose rather than the interval. Braking is DCT hypertrophy as natriuresis contracts the ECF. Loop toxicity: hypovolaemia, hypokalaemia, hypomagnesaemia. The prognostic value is the furosemide stress test. On the KDIGO reconciliation: the Mehta association between diuretics and mortality is confounding by indication, and it reverses once a negative balance is achieved (HR 0.25). 2 min. The uraemic inhibitors of OAT secretion (indoxyl sulfate, p-cresyl sulfate) and the albumin-binding competition are why the same plasma furosemide level under-delivers in AKI — this is mechanism, not ritual, for the higher dose. Braking is DCT hypertrophy/hyperplasia, reversible with thiazide; bioavailability differences explain the oral-vs-IV gap (oral through a congested gut is a lottery).

The furosemide stress test — a functional biomarker you already own

Dose: 1.0 mg/kg IV naive · 1.5 mg/kg IV prior loop
Replace urine mL for mL
2-h UO cut-off 200 mL
<200 mL = high risk of stage 3 or KRT.

Test performance

Chawla 2013: AUC 0.87 for stage-3 progression
Rewa 2019: sens 73.9%, spec 89.9%
Adding it to [TIMP-2]×[IGFBP7] improves discrimination

How to act on the result

Pass — de-escalate monitoring, do not rush access
Fail — plan access, review renally cleared drugs
Prognostic, not therapeutic — a fail is not an order for KRT

Pitfall — running the test on the wrong patient

Invalid and dangerous in a hypovolaemic or hypotensive patient. Optimise volume first.

Chawla LS et al. Crit Care. 2013;17:R207 · Rewa OG et al. J Crit Care. 2019;52:109–114 · Koyner JL et al. J Am Soc Nephrol. 2015;26:2023 · McMahon BA et al. Biomarkers. 2018;23:61 · Koyner Ch 12

Diuretic resistance — work the ladder to its end

Resistance or under-dose? Urine Na⁺ <50–70 mmol/L 2 h after a dose = the prescription failed, not the kidney.

Double to the ceiling — 40 → 80 → 160 → 200 mg.

Infusion after a bolus, 5–20 mg/h.

Correct what blunts response. ADVOR: acetazolamide 500 mg IV daily → decongestion 42.2% vs 30.5%.

Sequential nephron blockade — metolazone 2.5–10 mg or chlorothiazide 500 mg IV before the loop.

Albumin co-administration — clinically marginal.

If the nephron cannot excrete the load, UF is next; ≤1.75 mL/kg/h.

Mullens W et al. N Engl J Med. 2022;387:1185–1195 (ADVOR) · Trullàs JC et al. Eur Heart J. 2023;44:411–421 (CLOROTIC) · Hoorn EJ, Ellison DH. Am J Kidney Dis. 2017;69:136 · Mullens W et al. Eur J Heart Fail. 2019;21:137 · Kitsios GD et al. J Crit Care. 2014;29:253 · Koyner Ch 12, 40

DOSE, CARRESS-HF, ROSE-AHF — the decongestion evidence

DOSE · N Engl J Med 2011

308 ADHF, 2×2 factorial. Bolus vs continuous: no difference. High dose: more diuresis, more transient Cr rise.

CARRESS-HF · N Engl J Med 2012

UF vs stepped care in worsening renal function: the 96-h endpoint favoured stepped care, with more serious events on UF.

ROSE-AHF · JAMA 2013

Low-dose dopamine or nesiritide added to a diuretic — neither improved urine volume.

Key point

A Cr rise during effective decongestion is haemoconcentration, not injury, and predicts better outcomes. Stop for hypotension, not for a number.

Felker GM et al. N Engl J Med. 2011;364:797–805 · Bart BA et al. N Engl J Med. 2012;367:2296–2304 · Chen HH et al. JAMA. 2013;310:2533–2543 (ROSE-AHF) · Mullens W et al. J Am Coll Cardiol. 2009;53:589 · Koyner Ch 40

Norepinephrine: writing the actual prescription

The order

Working range 0.05–0.5 µg/kg/min
Titrate q3–5 min to the written MAP target — rewrite it daily
Start alongside fluid, not after: CENSER, CLOVERS
Peripheral in a proximal large vein is a short-term bridge
Extravasation: stop, aspirate; phentolamine 5–10 mg

Escalation, not dose creep

As noradrenaline rises, add vasopressin 0.03 U/min
Wean vasopressin 0.01 U/min q30–60 min; abrupt stop causes hypotension
Rising dose + falling MAP = a missed diagnosis

NEE = NE + epinephrine + phenylephrine/10 + dopamine/150 + vasopressin × 2.5
Above ~0.5 µg/kg/min mortality rises steeply — escalate the diagnosis.

Koyner Ch 13 (Table 13.1) · Permpikul C et al. Am J Respir Crit Care Med. 2019;199:1097 (CENSER) · Russell JA et al. N Engl J Med. 2008;358:877 (VASST) · Goradia S et al. J Crit Care. 2021;61:233 · Surviving Sepsis Campaign 2026, Crit Care Med. 2026;54:725–812

Second-line and rescue agents — dose and practical caveat

Agent	Dose	Prescribing point
Vasopressin	0.01–0.04 U/min	Catecholamine-sparing; constricts the efferent arteriole. Wean slowly.
Epinephrine	0.01–0.5 µg/kg/min	CAT: no advantage vs NE. Don't chase β₂ lactate
Phenylephrine	0.1–10 µg/kg/min	Pure α₁: reflex bradycardia, falling CO
Dopamine	Restricted use only	SOAP II: double the arrhythmia rate. No "renal dose"
Angiotensin II	20 ng/kg/min, titrate to 80 ng/kg/min	ATHOS-3: MAP response 69.9% vs 23.4%
Dobutamine	2.5–20 µg/kg/min	Can drop MAP; reserve for low CO with congestion
Milrinone	0.125–0.75 µg/kg/min	Renally cleared — accumulates in AKI

Koyner Ch 13 · Myburgh JA et al. Intensive Care Med. 2008;34:2226 (CAT) · De Backer D et al. N Engl J Med. 2010;362:779 · Khanna A et al. N Engl J Med. 2017;377:419 (ATHOS-3) · Tumlin JA et al. Crit Care Med. 2018;46:949

Terlipressin in HRS-AKI: real benefit, real respiratory cost

CONFIRM · N Engl J Med 2021

300 HRS-1 patients: verified reversal 32% vs 17%. No 90-day survival benefit — respiratory failure 11% vs 2%.

How to give it

1 mg IV q6h; 2 mg q6h if Cr has not fallen ≥25% by day 4
With albumin 20–40 g/day
Stop day 5–7 without response, or on KRT
Response predictors: ΔMAP >5 mmHg, Cr <5, bili <10 mg/dL

Before you write it

Screen oxygenation; avoid in hypoxaemia and ACLF 3
Unavailable? Norepinephrine + albumin works comparably
Ward alternative: midodrine + octreotide
Stop diuretics and NSAIDs — still a diagnosis of exclusion

Wong F et al. N Engl J Med. 2021;384:818–828 (CONFIRM) · Angeli P et al. J Hepatol. 2019;71:811 (ICA consensus) · Koyner Ch 42 (Table 42.2)

Blood, haemostasis, nutrition & glycaemia

Four prescriptions the consultant nephrologist is expected to have an opinion about — and usually does not.

Koyner Ch 11, 14–15

Transfusion: 7 g/dL everywhere, except where it is not

Trial	Population	Restrictive vs liberal
TRICC · 1999	838 ICU	30-d mortality 18.7% vs 23.3%
TRISS · 2014	998 septic shock	90-d mortality 43.0% vs 45.0%
TRICS III · 2017	5,243 cardiac surgery	Restrictive non-inferior; AKI identical
MINT · 2023	3,504 MI	30-d death/MI/revasc 16.9% vs 14.5%

Transfuse above 7

ACS or ongoing myocardial ischaemia.

Transfuse to physiology

Major haemorrhage — treat the bleeding rate, not the lagging Hb.

Transfuse less

Potential transplant candidates — minimise HLA exposure.

Pitfall — TACO, the transfusion reaction nephrologists cause

The commonest cause of transfusion-related death; unlike TRALI, it responds to a diuretic
One unit slowly, then recheck

Hébert PC et al. N Engl J Med. 1999;340:409 · Holst LB et al. N Engl J Med. 2014;371:1381 · Mazer CD et al. N Engl J Med. 2017;377:2133 · Carson JL et al. N Engl J Med. 2023;389:2446 (MINT) · Holcomb JB et al. JAMA. 2015;313:471 (PROPPR) · Koyner Ch 11 (Table 11.2), Ch 47

Haemostasis in kidney failure: thresholds, the uraemic toolkit, anticoagulants

Product	Threshold	Nephrology note
Platelets — prophylaxis	≤10 × 10⁹/L	No evidence above this
Platelets — central line	>20 × 10⁹/L	Tunnelled catheters count
Platelets — LP	≥50 × 10⁹/L	Not required before bypass
Platelets — kidney biopsy	≥100 × 10⁹/L	Custom, not evidence
Fresh-frozen plasma	Massive transfusion	Highest TRALI risk
Cryoprecipitate	Fibrinogen <100 with bleeding	Fibrinogen, VIII
Prothrombin complex concentrate	Warfarin reversal	Faster than FFP

Anticoagulation traps in AKI

HIT — platelets >50% down day 5–10: stop heparin; argatroban
DOAC reversal — idarucizumab; andexanet; PCC fallback
LMWH — accumulates below eGFR 30: halve it or use UFH

Hedges SJ et al. Nat Clin Pract Nephrol. 2007;3:138–153 · AABB platelet guideline 2015 · Koyner Ch 11, Ch 14 (Table 14.1) · citrate toxicity from massive transfusion causes ionised hypocalcaemia and QT prolongation — same chemistry as regional citrate anticoagulation (Lecture 9)

The uraemic bleeding toolkit — a ladder

Correct the correctable — stop antiplatelets; treat anaemia; correct acidaemia and ionised hypocalcaemia.

Dialyse — the only cause-directed therapy. Everything below is a bridge.

Desmopressin 0.3 µg/kg IV — onset ~1 h, tachyphylaxis after 2 doses.

Cryoprecipitate, 10 units — reserve for active bleeding.

Conjugated oestrogens 0.6 mg/kg/day × 5 days — for angiodysplastic GI bleeding.

Tranexamic acid — renally cleared; dose-reduce.

Hedges SJ et al. Nat Clin Pract Nephrol. 2007;3:138–153 · AABB platelet guideline 2015 · Koyner Ch 11

Feeding the AKI patient — and the glucose target

Target	Prescription	Why
Energy	20–25 kcal/kg/day	Set by the illness, not by AKI
Protein — no KRT	0.8–1.2 g/kg/day	A high load during active injury may aggravate it
Protein — on KRT	1.2–1.5 g/kg/day, up to 1.7	Includes ~0.2 g/kg/day for losses
Route	Enteral first, advanced slowly	AKI impairs GI motility
Micronutrients	Water-soluble vitamins 2× RDA	CKRT strips vitamins and trace elements
Glucose	140–180 mg/dL	NICE-SUGAR: tight control raised mortality

Two rules and one trial to remember

Never restrict protein to postpone dialysis
Count the hidden calories: citrate ≈200–600 kcal/day; propofol is lipid

Koyner Ch 15 (Tables 15.1, 15.3–15.4) · Fiaccadori E et al. Clin Nutr. 2021 (ESPEN, kidney disease) · NICE-SUGAR Investigators. N Engl J Med. 2009;360:1283–1297 · Jaber S et al. Lancet. 2018;392:31–40 · KDIGO AKI 2012

Cases, polls & wrap-up

Four consults where the correct answer is usually to stop doing something — and two polls where a single number decides the plan.

Koyner Ch 10, 12, 36, 40, 42–43

Case 1 · 64 M, urosepsis, hour 4 — "one more litre because he's still oliguric"

Obstructive pyelonephritis, stented; 30 mL/kg saline in the ED.

MAP 62NE 0.18 µg/kg/minLactate 3.8Cr 1.0 → 2.1UO 0.2 mL/kg/hPLR ΔSV +4%Cl⁻ 112HCO₃⁻ 18AG 11

Stop the volumePLR ΔSV +4% is negative. The next litre buys nothing.

→

Fix the pressureNE to MAP 65; add vasopressin as NE rises.

→

Name the acidosisNormal gap at Cl⁻ 112 — switch to balanced.

→

Leave the urine aloneEvolving septic ATN. No boluses, no furosemide.

Key point

"Still oliguric" is not an indication.

Koyner Ch 10, 12, 36 · Semler MW et al. N Engl J Med. 2018 · Shapiro NI et al. N Engl J Med. 2023 · Surviving Sepsis Campaign 2026, Crit Care Med. 2026;54:725–812

Case 2 · 74 F, HFrEF (EF 30%), day 5 — resistance, or a bad prescription?

Home furosemide 80 mg bd, now 80 mg IV bd. Gained 7 kg.

Cr 1.4 → 2.3K⁺ 3.2HCO₃⁻ 34Cl⁻ 91CVP 18Urine Na⁺ 12+6 L

Urine Na⁺ 12 after a loop dose = failed prescription
On 160 mg/day — DOSE high-dose was ≈ 400 mg/day
HCO₃⁻ 34, Cl⁻ 91 — hypochloraemic alkalosis
A Cr rise at CVP 18 is congestive

What not to do

No ultrafiltration first
Don't stop the diuretic for the Cr
No albumin adjuvant

The three orders

1. Furosemide 160 mg IV → infusion. 2. Acetazolamide 500 mg IV. 3. Chlorothiazide 500 mg.

Mullens W et al. N Engl J Med. 2022;387:1185 (ADVOR) · Trullàs JC et al. Eur Heart J. 2023;44:411 (CLOROTIC) · Felker GM et al. N Engl J Med. 2011;364:797 · Bart BA et al. N Engl J Med. 2012;367:2296 · Koyner Ch 12, 40

Case 3 · 58 M, alcohol-related cirrhosis — HRS-AKI, or something cheaper to fix?

Tense ascites; 8 L paracentesis 3 days ago, no albumin. On furosemide, spironolactone, carvedilol.

MELD 28Cr 0.9 → 2.8Na⁺ 126MAP 66Urine Na⁺ 6Proteinuria 0.2 g/dayBland sedimentNormal ultrasoundAscites PMN 90Bili 8.4

The fork: post-paracentesis dysfunction and diuretic hypovolaemia are reversible mimics
Step 1: stop diuretics and β-blocker; albumin 1 g/kg/day × 2 d
Step 2: vasoconstrictor + albumin — terlipressin 1 mg q6h
PMN 90 excludes SBP

Two errors to avoid

Skipping albumin for >5 L paracentesis — likely the cause here
Terlipressin without screening oxygenation

Angeli P et al. J Hepatol. 2019;71:811 (ICA) · Wong F et al. N Engl J Med. 2021;384:818 (CONFIRM) · Sort P et al. N Engl J Med. 1999;341:403 · China L et al. N Engl J Med. 2021;384:808 (ATTIRE) · Koyner Ch 42

Case 4 · 52 M, severe acute pancreatitis, day 3 — the trap in the numbers

Ventilated, sedated; 11 L crystalloid. Tense abdomen. Surgery asks for CRRT; the registrar wants a bolus.

Bladder 26 mmHgMAP 70 → APP 44UO 8 mL/hCr 1.1 → 2.9PPV 16%NE 0.22 µg/kg/min+9 L

This is ACS: IAP >20 with organ failure
The trap: IAH compresses the IVC and fabricates PPV
Target APP = MAP − IAP ≥ 60
Medical bundle: sedation ± paralysis; decompression; negative balance
Refractory → decompressive laparotomy

What CRRT is and is not for here

UF can lower IAP — but cannot decompress ACS
KRT "for the oliguria" treats the alarm, not the compartment

Koyner Ch 43 · Kirkpatrick AW et al. Intensive Care Med. 2013;39:1190–1206 (WSACS) · Cheatham ML et al. J Trauma. 2000;49:621 · Malbrain MLNG et al. Minerva Anestesiol. 2014;80:293

Poll 1 · The furosemide stress test

Day 3 of KDIGO stage 2 AKI after pancreatitis. Neutral balance, MAP 74 off vasopressors, CVP 9, no obstruction. She had furosemide two days ago. You give furosemide 1.5 mg/kg IV and replace urine millilitre for millilitre. Two-hour output: 130 mL.

A. Wrong dose for a non-naive patient — repeat at 2 mg/kg · B. High risk of stage 3 or KRT — plan access, review drug dosing · C. Start an infusion to force the output up · D. Give 500 mL and repeat the test · E. Uninterpretable — send [TIMP-2]×[IGFBP7]

Hands up — then say what you would change in the next 24 hours

Chawla LS et al. Crit Care. 2013;17:R207 · Rewa OG et al. J Crit Care. 2019;52:109–114 · Koyner JL et al. J Am Soc Nephrol. 2015;26:2023 · Koyner Ch 10, 12

Poll 2 · Choosing the fluid when the trials are neutral

In which patient is Plasma-Lyte 148 instead of 0.9% saline most likely to cause net harm?

A. Septic shock, K⁺ 5.8, oligoanuric stage 3 AKI · B. Severe traumatic brain injury, Na⁺ 139, ICP monitor · C. Decompensated cirrhosis, lactate 3.4 · D. DKA, pH 7.05, chloride 118 · E. Post-operative, 2 L/day nasogastric aspirate, HCO₃⁻ 36

Hands up — then defend the answer you rejected most confidently

Zampieri FG et al. JAMA. 2021;326:818–829 · Hammond NE et al. Lancet Respir Med. 2024;12:29–38 (BEST-Living) · Koyner Ch 8, 10

Answer B. The BaSICS pre-specified TBI subgroup favoured saline, and the BEST-Living individual-patient-data meta-analysis (6 trials, 34,450 patients) found the probability of benefit exceeded 0.97 in patients without TBI while the direction reversed in TBI. The mechanism is tonicity, not buffer: Plasma-Lyte is 295 mOsm/L and Ringer's only ~273, and lactate is not an effective osmole — a functionally hypotonic fluid worsens cerebral oedema. A is a myth: balanced solutions contain 4–5 mmol/L K⁺, but saline's hyperchloraemic acidosis shifts more potassium out of cells than that; balanced fluids lower serum K⁺ on average. C is the tempting distractor — impaired hepatic lactate metabolism argues against Ringer's LACTATE, but Plasma-Lyte is buffered with acetate and gluconate, metabolised largely in muscle, so it is the right choice in liver failure. D is wrong in the other direction: DKA with hyperchloraemia is a setting where balanced fluid speeds resolution of acidosis. E is a genuine saline indication (chloride-depletion alkalosis) but omitting saline there causes a slower correction, not net harm, and acetazolamide addresses it too. Teaching point: for the fluid choice, the exception that actually kills is tonicity in the injured brain. 60 s.

Key takeaways

Balanced by default, saline by indication — and the exception that matters is TBI.
Albumin earns its cost in SBP, large-volume paracentesis and HRS-AKI; to a serum target it failed twice.
The exact volume rarely decides survival — but >10% accumulation predicts death.
Diuretics give volume control, never renal protection. The FST is the best functional biomarker you own.
For resistance: measure natriuresis, double to the ceiling, acetazolamide, then block the distal tubule.
Transfuse at 7 g/dL except acute MI; glucose 140–180; protein 1.2–1.5 g/kg/day on KRT.

References & further reading

Where to go deeper

Semler MW, et al. SMART. NEJM. 2018;378:829.
Zampieri FG, et al. BaSICS. JAMA. 2021;326:818.
China L, et al. ATTIRE. NEJM. 2021;384:808.
Meyhoff TS, et al. CLASSIC. NEJM. 2022;386:2459.
Chawla LS, et al. FST. Crit Care. 2013;17:R207.
Mullens W, et al. ADVOR. NEJM. 2022;387:1185.
Khanna A, et al. ATHOS-3. NEJM. 2017;377:419.
Carson JL, et al. MINT. NEJM. 2023;389:2446.

Full list: (1) SMART (N Engl J Med 2018;378:829–839) + SALT-ED (819–828); (2) BaSICS (JAMA 2021;326:818–829), PLUS (N Engl J Med 2022;386:815–826), BEST-Living IPD meta-analysis (Lancet Respir Med 2024;12:29–38); (3) ATTIRE (N Engl J Med 2021;384:808–817), ALBIOS (2014;370:1412–1421), Sort (1999;341:403–409); (4) CLASSIC (2022;386:2459–2470), CLOVERS (2023;388:499–510), PICARD (Kidney Int 2009;76:422–427); (5) Chawla FST (Crit Care 2013;17:R207) + Rewa (J Crit Care 2019;52:109–114); (6) ADVOR (2022;387:1185–1195), CLOROTIC (Eur Heart J 2023;44:411–421), DOSE (2011;364:797–805), CARRESS-HF (2012;367:2296–2304), ROSE-AHF (JAMA 2013;310:2533–2543); (7) SEPSISPAM (2014;370:1583–1593), ATHOS-3 (2017;377:419–430), CONFIRM (2021;384:818–828); (8) TRICC (1999;340:409–417), TRISS (2014;371:1381–1391), MINT (2023;389:2446–2456), NICE-SUGAR (2009;360:1283–1297); (9) WSACS (Intensive Care Med 2013;39:1190–1206); (10) KDIGO AKI 2012 (Kidney Int Suppl;2:1–138) + SSC 2026 (Crit Care Med 2026;54:725–812, updating SSC 2021); (11) Koyner Handbook of Critical Care Nephrology, Wolters Kluwer 2021, Ch 8–15, 40, 42–43, 46–47.

Critical Care Nephrology · Two-Week Intensive

Thank you

Questions & discussion — bring a fluid balance chart from your own unit and we will price it.

Next: Lecture 04 — Biomarkers, Risk Scores & Imaging in AKI