Critical Care Nephrology · Two-Week Intensive · Lecture 1 of 9

AKI Essentials

Definitions, staging, epidemiology & etiologies of acute kidney injury in the ICU

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…

  1. Stage AKI with KDIGO 2012 criteria — and name three situations where the criteria mislead you.
  2. Place a patient on the AKI → AKD → CKD continuum and distinguish transient from persistent AKI.
  3. Build a systematic prerenal / intrinsic / postrenal differential for the multifactorial ICU patient.
  4. Execute a first-24-hours workup: history, sediment, chemistries, ultrasound — and know when to biopsy.
  5. Quote the epidemiology that justifies urgency to the ICU team.

Koyner Ch 5–7 · KDIGO 2012 AKI Guideline · Chawla LS et al. Nat Rev Nephrol 2017

01

Why AKI matters

The most common organ dysfunction we are consulted for — and the one most often called "expected".

Koyner Ch 5–6 · Hoste 2015 · Susantitaphong 2013

How common is AKI — and where?

57%of ICU patients meet KDIGO criteria (AKI-EPI, 97 centres)
1 in 5of all hospitalised adults worldwide (pooled 21.6%)
13.5%of ICU patients receive kidney replacement therapy
↑ 2–3×rising incidence over two decades: age, comorbidity, detection
Frame it this way

In a 20-bed ICU, roughly 11 patients will have AKI today and 2–3 will need dialysis before discharge.

Hoste EAJ et al. Intensive Care Med 2015;41:1411 (AKI-EPI) · Susantitaphong P et al. Clin J Am Soc Nephrol 2013;8:1482 · Koyner Ch 6

AKI is a driver of poor outcomes, not just a marker

In hospital

  • Mortality rises stepwise with KDIGO stage
  • AKI needing KRT: 90-day mortality ≈ 45–50%
  • More ventilator days, more fluid overload, more infection

Distant organ injury

  • Lung: capillary leak, impaired fluid clearance
  • Heart: arrhythmia, myocardial depression
  • Brain: uremic encephalopathy, delirium

After discharge

  • New/progressive CKD (pooled HR ≈ 8.8)
  • ESKD (HR ≈ 3.1), death (HR ≈ 2.6)
  • Excess risk persists in "recovered" patients

Coca SG et al. Kidney Int 2009;76:893 · See EJ et al. Kidney Int 2019;95:160 · Koyner Ch 6, 51

“AKI is a clinical syndrome defined by laboratory and urine-output criteria — not a single disease with a single cause. The criteria tell you that the kidney is injured; they never tell you why.”

— The mental frame for every ICU consult

02

Definitions & staging

Twenty years of consensus criteria — powerful, standardised, and built on two imperfect signals.

Koyner Ch 5 · KDIGO 2012 · Levey Kidney Int 2020

Two decades of consensus criteria

YearSystemKey innovationLegacy problem
2004RIFLEFirst consensus: Risk · Injury · Failure · Loss · ESKD; introduced urine outputNeeded a baseline creatinine
2007AKINAdded the 48-h absolute rise ≥0.3 mg/dL; 3 stages; no eGFR criterionOver-called small fluctuations
2012KDIGOMerged both: 48-h absolute or 7-day relative rise — still the standardCreatinine and urine output are both indirect
2017–21ADQI / KDIGO nomenclatureAdded AKD, transient vs persistent AKI, sub-phenotypesNot yet in most EHR alerts

Bellomo R et al. Crit Care 2004 · Mehta RL et al. Crit Care 2007 · KDIGO AKI Work Group, Kidney Int Suppl 2012 · Chawla LS et al. Nat Rev Nephrol 2017

KDIGO staging — the language of the ICU

StageSerum creatinineUrine output
11.5–1.9× baseline (within 7 d) or ≥0.3 mg/dL rise within 48 h<0.5 mL/kg/h × 6–12 h
22.0–2.9× baseline<0.5 mL/kg/h × ≥12 h
3≥3.0× baseline, or rise to ≥4.0 mg/dL, or start of KRT, or eGFR <35 mL/min/1.73 m² (age <18 y)<0.3 mL/kg/h × ≥24 h or anuria ≥12 h
Key point

Either criterion can stage, and the higher of the two wins. Document both — and document the baseline you used.

KDIGO AKI Work Group, Kidney Int Suppl 2012;2:19–36

Staging in practice: a worked example

Bedside numbers
Weight 70 kgBaseline Cr 0.9Day 1 Cr 1.3Day 2 Cr 2.1UO 190 mL / 12 h
Cr

2.1 ÷ 0.9 = 2.3× baseline → creatinine stage 2

UO

190 mL ÷ 70 kg ÷ 12 h = 0.23 mL/kg/h for 12 h → urine-output stage 2 (needs ≥24 h for stage 3)

Final: KDIGO stage 2 AKI, both axes concordant — and already 48 h into the injury

Three habits that prevent errors

  • Use ideal or adjusted body weight in the obese — actual weight falsely reassures
  • Read UO in blocks (6/12/24 h), never a single hour
  • State the baseline explicitly in the note: "baseline Cr 0.9 from clinic, 4 months ago"

KDIGO 2012 · Koyner Ch 5

Urine output: powerful, but easy to get wrong

Why it earns its place

  • Detects injury hours to days earlier than creatinine
  • Oliguria alone (normal Cr) still predicts mortality and KRT
  • Free, continuous, already being measured

Why it fails

  • Diuretics uncouple flow from function — UO looks "fine"
  • Needs a catheter and honest hourly charting
  • Obesity: mL/kg/h on actual weight under-calls oliguria
  • Non-oliguric ATN is common — normal UO never excludes AKI
Common pitfall

"Creatinine is fine" does not rule out AKI when the urine-output criterion is met — and a patient on a furosemide infusion can be anuric-in-truth while charting 60 mL/h.

KDIGO 2012 · Kellum JA et al. J Am Soc Nephrol 2015;26:2231 · Koyner Ch 5

The baseline creatinine problem

KDIGO needs a baseline. In the ICU it is often unknown — and the choice changes the stage, which changes the decisions.

Best
Stable outpatient value 7–365 days before admission
Acceptable
Lowest inpatient creatinine during this admission
Last resort
Back-calculate assuming eGFR 75 mL/min/1.73 m² (MDRD) — systematically over-diagnoses AKI in CKD
Never
Admission creatinine in a patient who arrived already injured

Why it matters

A patient with true baseline Cr 1.8 (CKD 3b) back-calculated to 0.9 is instantly mislabelled stage 2 — triggering nephrotoxin stops, imaging and sometimes a dialysis conversation that was never needed.

KDIGO 2012 · Siew ED et al. Kidney Int 2010;77:536 · Koyner Ch 5

Creatinine is a lagging, diluted, muscle-dependent signal

Lag

Creatinine rises only after GFR has already fallen substantially, and takes 24–72 h to reach a new steady state. Early injury looks normal.

Dilution

Aggressive resuscitation expands total body water and dilutes creatinine — the same kidney looks "better" after 5 L of fluid.

Production

Sarcopenia, liver failure, immobility and sepsis all lower creatinine generation. A "normal" 0.6 mg/dL in an ICU patient may mean GFR 40.

Kinetic eGFR

Uses the rate of change of creatinine rather than a steady-state value — estimates real-time GFR during rapidly changing injury.

Cystatin C

Independent of muscle mass; the combined creatinine–cystatin C CKD-EPI 2021 equation is the most accurate and is race-free. Confounded by steroids and thyroid disease.

Chen S. J Am Soc Nephrol 2013;24:877 · Inker LA et al. N Engl J Med 2021;385:1737 · Koyner Ch 5, 16 (biomarkers in Lecture 4)

AKI is not a one-off event: the continuum

0–48 H

Transient AKIReverses within 48 h — usually haemodynamic

2–7 D

Persistent AKICriteria still met beyond 48 h — worse outcomes

7–90 D

AKDAcute kidney disease — recovery or progression is decided here

>90 D

CKDResidual damage — lifelong follow-up

Key point

Every AKI consult is also a referral for future kidney care. The AKD window (7–90 days) is where nephrology follow-up actually changes trajectories — and it is the window most often dropped at discharge.

Chawla LS et al. Nat Rev Nephrol 2017;13:241 (ADQI 16) · Levey AS et al. Kidney Int 2020;97:1117 · Koyner Ch 51

Sub-phenotypes: not all "stage 2" is the same patient

Transient / rapidly reversing

  • Resolves ≤48–72 h after perfusion is restored
  • Usually haemodynamic or volume-responsive
  • Mortality close to no-AKI patients

Persistent

  • Criteria met >48 h despite correction of haemodynamics
  • Structural injury; biomarker-positive
  • Independently predicts KRT, non-recovery and death

Renal angina

Combine risk (sepsis, transplant, ventilation, pressors) with early injury signs (fluid overload, small Cr change) to decide who deserves a biomarker or a closer look — the "chest pain" analogy for the kidney.

Why you should care

Trial enrolment, biomarker use and KRT-timing decisions all now hinge on this distinction. "Will it reverse?" is a more useful question than "what stage?"

Chawla LS et al. Nat Rev Nephrol 2017 (ADQI 16) · Goldstein SL, Chawla LS. Clin J Am Soc Nephrol 2010;5:943 · Koyner Ch 5, 7

03

Etiologies

The three-compartment framework still works — as long as you remember ICU patients rarely have only one cause.

Koyner Ch 5, 8–9 · Ronco Ch 11–16

The classic framework: three compartments

Prerenal ~30%

Hypoperfusion without structural injury — responds to restoring perfusion, by definition only diagnosable in retrospect.

Intrinsic ~60%

Damage to tubules, interstitium, glomeruli or vessels — the ICU majority, and mostly acute tubular injury.

Postrenal <10%

Obstruction — small in numbers, large in consequences if missed, and the only fully reversible cause.

Common pitfall

ICU AKI is usually multifactorial: sepsis plus hypovolemia plus nephrotoxins plus raised intra-abdominal pressure. Naming one cause and stopping the search is the commonest consult error.

Koyner Ch 5 · Uchino S et al. JAMA 2005;294:813 (BEST Kidney: sepsis in ~50% of severe ICU AKI)

Prerenal AKI — the reversible one

Mechanisms

  • True hypovolemia: haemorrhage, GI losses, burns, diuresis
  • Low cardiac output: cardiogenic shock, tamponade, PE
  • Vasodilation: sepsis, cirrhosis, anaphylaxis, anaesthesia
  • Efferent arteriolar failure: ACEi/ARB, NSAIDs (afferent), calcineurin inhibitors
  • Venous congestion: right heart failure, raised IAP, fluid overload

Signatures (before diuretics)

  • BUN/Cr >20:1, FeNa <1%, FeUrea <35%
  • UOsm >500 mOsm/kg, urine Na <20 mEq/L
  • Bland sediment, hyaline casts only
  • Creatinine falls within 24–72 h of correcting the insult
Key point

Prerenal and ATN are the two ends of one spectrum, not two diseases. Prolonged or repeated hypoperfusion converts one into the other — and the transition has no laboratory marker.

Koyner Ch 5 · Ronco Ch 18

Sepsis-associated AKI: the model that broke the model

Septic AKI is not simply renal ischaemia. Renal blood flow is often normal or increased, and biopsies show remarkably little necrosis.

  • Microcirculatory failure — heterogeneous peritubular flow despite adequate global flow
  • Inflammation — PAMPs/DAMPs signalling through tubular TLRs; leukocyte adhesion
  • Adaptive metabolic downregulation — tubular cells switch off energy-costly transport to survive; a defence, not death
  • Venous congestion — the right-sided contribution nobody measures

The clinical consequence

More MAP and more fluid do not automatically mean more filtration. Once perfusion is adequate, additional fluid is congestive, not restorative — this is the physiology behind Lecture 2 and 3.

Takasu O et al. Am J Respir Crit Care Med 2013;187:509 · Gomez H, Kellum JA. Crit Care Clin 2016 · Langenberg C, Bellomo R. Crit Care 2005 · Ronco Ch 45

Intrinsic AKI — acute tubular injury (ATN)

Ischemic ATN

  • Shock of any type, cardiac arrest, prolonged clamping, CPB
  • Outer medulla (S3 segment, thick ascending limb) is the watershed
  • Muddy-brown granular casts, renal tubular epithelial cells

Nephrotoxic ATN

  • Aminoglycosides (5–10 d lag, non-oliguric, Mg/K wasting)
  • Vancomycin, amphotericin B (deoxycholate ≫ liposomal), cisplatin, tenofovir
  • Crystal: acyclovir, methotrexate, sulfadiazine — need volume and often alkalinisation
  • Pigment: rhabdomyolysis, intravascular haemolysis (Lecture 6)
Key point

Most ICU AKI is tubular injury on a septic + ischaemic + nephrotoxic background. You can only modify the last two — so do.

Koyner Ch 5, 8–9 · Perazella MA. Clin J Am Soc Nephrol 2019;14:1101

Intrinsic AKI — beyond the tubule

SiteDiseaseClue at the bedsideDo not miss
InterstitiumAcute interstitial nephritisNew drug (PPI, β-lactam, NSAID, checkpoint inhibitor), fever, rash, sterile pyuria, WBC castsStop the drug; steroids if no recovery in 5–7 d
GlomerulusAcute GN / RPGN, endocarditis-associatedHematuria, dysmorphic RBCs, RBC casts, proteinuria, hypertension, low complementSerologies today; this is the biopsy group
Small vesselsTMA: HUS/TTP, drug-induced, malignant hypertension, scleroderma renal crisisSchistocytes, thrombocytopenia, LDH ↑, haptoglobin ↓, normal coagulationADAMTS13 before plasma exchange; eculizumab in aHUS
Large vesselsRenal artery thrombo-embolism, atheroembolic disease, renal vein thrombosisSudden anuria after instrumentation, livedo reticularis, blue toes, eosinophilia, LDH ↑Atheroemboli after any aortic procedure — subacute, stepwise decline

Koyner Ch 5 · Ronco Ch 47 · Perazella MA, Rosner MH. Clin J Am Soc Nephrol 2022

Postrenal AKI — always rule out obstruction

Where it comes from

  • Prostate, pelvic or retroperitoneal malignancy, retroperitoneal fibrosis
  • Stones (bilateral, or single functioning kidney), papillary necrosis
  • Clot retention, neurogenic bladder, blocked or kinked catheter
  • Intratubular: crystals, myeloma cast nephropathy

How it presents

  • Anuria alternating with polyuria is classic but uncommon
  • Partial obstruction can present with normal urine output
  • Post-obstructive diuresis after relief — replace, but do not chase
Common pitfall

A Foley in place is not proof of patency — flush it. And early or volume-depleted obstruction can show no hydronephrosis on ultrasound: if the story fits, image again or get a CT.

Koyner Ch 5, 18

Nephrotoxins — the part of the differential you can actually change

Antimicrobials

Vancomycin (AUC-guided dosing), aminoglycosides, amphotericin B, acyclovir, colistin, sulfonamides, tenofovir

Haemodynamic agents

NSAIDs, ACEi/ARB, calcineurin inhibitors, diuretic excess, iodinated contrast, SGLT2 inhibitors (volume, euglycemic DKA)

ICU staples

Hydroxyethyl starch (avoid entirely), high-dose IVIG (sucrose), mannitol, chronic high-dose loop infusions, herbal/traditional preparations

Nuance worth knowing

The vancomycin + piperacillin-tazobactam AKI signal comes from observational data. The randomised ACORN trial (cefepime vs piperacillin-tazobactam, 2023) found no difference in AKI — much of the signal is probably creatinine-secretion interference, not injury. Do not let it drive antibiotic choice alone.

Koyner Ch 5, 8 · Qian ET et al. JAMA 2023;330:1557 (ACORN) · Perazella MA. Clin J Am Soc Nephrol 2019

Organ crosstalk syndromes — when the kidney is the messenger

Cardiorenal syndrome (Ronco types 1–5)

  • Type 1 acute cardiac → renal · Type 2 chronic HF → CKD
  • Type 3 acute renal → cardiac · Type 4 CKD → cardiac
  • Type 5 systemic (sepsis, amyloid) → both
  • Congestion, not low output, is the dominant driver in most type-1 patients

Hepatorenal syndrome — AKI (HRS-AKI)

  • ICA 2019: diagnose by KDIGO criteria — the old fixed Cr ≥2.5 mg/dL threshold is gone
  • Requires: cirrhosis + ascites, no shock, no nephrotoxins, no response to albumin 1 g/kg × 2 d, no structural disease
  • Treat with terlipressin + albumin (CONFIRM, NEJM 2021) — watch for respiratory failure

Ronco C et al. J Am Coll Cardiol 2008;52:1527 · Angeli P et al. J Hepatol 2019;71:811 (ICA) · Wong F et al. N Engl J Med 2021;384:818 · Koyner Ch 40

04

The first 24 hours of the consult

A reproducible sequence: exposure history, the bedside, the urine, the ultrasound — then a decision about biopsy.

Koyner Ch 5, 18

A consult checklist that never fails

History & exposure

  • Volume losses: diarrhoea, bleeding, burns, drains, third spacing
  • Every hypotensive episode — read the flowsheet, not the summary
  • Contrast, surgery, CPB time, IABP/ECMO cannulation
  • Full medication reconciliation including pre-admission NSAIDs and herbals
  • Baseline kidney function — ask, then verify against a real laboratory value

Exam & bedside data

  • MAP trend and perfusion pressure (MAP − CVP), rhythm, pressor dose
  • JVP, peripheral oedema, capillary refill, mottling score
  • Intra-abdominal pressure in every unstable or distended patient
  • Bladder scan and catheter flush — takes two minutes
  • Skin: rash, purpura, livedo; eyes and joints for systemic disease

Koyner Ch 5 · KDIGO 2012 (Section 2: risk assessment)

Urine: the cheapest biopsy in medicine

Sediment — look yourself

  • Muddy-brown granular casts + tubular cells → ATN (a cast score predicts severity)
  • Dysmorphic RBCs / RBC casts → glomerular disease
  • WBC casts, sterile pyuria → interstitial nephritis (urine eosinophils are not reliable)
  • Crystals: uric acid, oxalate (ethylene glycol), acyclovir needles

Chemistry — collect early

  • FeNa and FeUrea are only interpretable before diuretics
  • UOsm, urine Na/K/Cl, spot protein-to-creatinine ratio
  • Urine dipstick blood positive but no RBCs → myoglobin or haemoglobin
  • Urine anion gap when a non-gap acidosis appears (Lecture 7)

Koyner Ch 5 · Perazella MA et al. Clin J Am Soc Nephrol 2010;5:402 (cast scoring)

Urine chemistries: prerenal vs ATN — and their limits

IndexPrerenalATNCaveat that ruins it
FeNa<1%>2%Useless after any diuretic; low in contrast AKI, pigment AKI, HRS and early GN
FeUrea<35%>50%Better after diuretics, but low in HRS and with urea recirculation
UOsm>500<350Needs concentrating ability — meaningless in CKD, DI or osmotic diuresis
Urine Na<20>40Tracks sodium intake and diuretics more than tubular health
BUN/Cr>20:110–15:1Raised by GI bleeding, steroids, catabolism; lowered by liver disease, malnutrition
Key point

These indices are supporting evidence, never a verdict. In the right patient, a carefully monitored fluid challenge with a dynamic endpoint answers the question faster than any ratio.

Koyner Ch 5 · Ronco Ch 55

Renal ultrasound — fast, cheap, bedside

  • Hydronephrosis → obstruction; false negatives in early, volume-deplete and encasing retroperitoneal disease
  • Size & echogenicity — kidneys <9 cm with thin cortex mean this is not a new problem
  • Bladder — the two most useful minutes of the study
  • Doppler resistive index — RI >0.75 suggests persistent AKI, but it is confounded by age, heart rate, pulse pressure and pressors
  • Renal vein flow pattern — the congestion signal (VExUS, Lecture 4)

What ultrasound cannot do

It cannot distinguish prerenal from ATN, cannot exclude early obstruction, and cannot diagnose glomerular disease. Order it to answer a specific question, not as a reflex.

Koyner Ch 18 · Ronco Ch 33 · Darmon M et al. Intensive Care Med 2011;37:68 (RI)

Putting it together: prerenal vs ATN

FeaturePrerenalAcute tubular injury
Response to volumeCreatinine falls in 24–72 hNo immediate response; may worsen with more fluid
Urine chemistryFeNa <1%, UOsm >500, BUN/Cr >20FeNa >2%, UOsm <350, BUN/Cr 10–15
SedimentBland, hyaline castsMuddy-brown granular casts, tubular cells
Damage biomarkersNegativePositive (NGAL, [TIMP-2]·[IGFBP7]) — Lecture 4
CourseDays1–3 weeks; a third will need KRT

Koyner Ch 5, 16

When does an ICU patient need a kidney biopsy?

Biopsy changes management when…

  • Active urinary sediment: RBC casts, dysmorphic haematuria, heavy proteinuria
  • Suspected RPGN, vasculitis, anti-GBM — days matter
  • AKI without a plausible haemodynamic or toxic explanation
  • No recovery beyond 3–4 weeks with an unclear cause
  • Suspected AIN where the culprit drug cannot be stopped empirically

Weigh against

  • Uncorrectable coagulopathy or platelets <50 ×10⁹/L
  • Uncontrolled hypertension, single kidney, hydronephrosis, active infection
  • Clinically significant bleeding ~1–2%; transfusion or intervention rarer
  • The transjugular route is an option when bleeding risk is high or the patient is ventilated and cannot be positioned
Key point

Ask one question before booking: "If the biopsy shows what I suspect, will I start immunosuppression tonight?" If the answer is no, the biopsy can usually wait.

Koyner Ch 5 · Ronco Ch 56 · Corapi KM et al. Am J Kidney Dis 2012;60:62

05

Epidemiology & outcomes

The numbers to quote when the team says "the creatinine is only a little up".

Koyner Ch 6 · Hoste 2015 · Uchino 2005

Risk factors — separate what you can act on

Susceptibility (fixed)

  • Age, baseline CKD, proteinuria, single kidney
  • Diabetes, hypertension, heart failure, cirrhosis, malignancy
  • Severity of illness, emergency vs elective surgery
  • Genetic: APOL1 risk variants in patients of African ancestry

Exposure (modifiable — today)

  • Hypotension: cumulative time below MAP 65 predicts AKI dose-dependently
  • Nephrotoxin burden — count them, stop what you can
  • Fluid overload and raised intra-abdominal pressure
  • Hyperglycaemia, anaemia, hypoxaemia
  • Timing of contrast and elective procedures
Key point

KDIGO risk stratification exists so that the bundle (stop nephrotoxins, optimise haemodynamics, monitor creatinine and UO, avoid hyperglycaemia) can be applied before stage 3 arrives — see Lecture 3.

Koyner Ch 6, 8 · Walsh M et al. Anesthesiology 2013;119:507 (MAP thresholds) · KDIGO 2012

Dose–response: stage and duration both predict death

Stage 1already carries measurable excess mortality
Stage 3mortality several-fold higher, independent of severity score
>7 dayspersistent AKI adds risk on top of peak stage
  • Severity and duration are independent — a brief stage 3 may fare better than a fortnight of stage 2
  • Repeat episodes compound the risk of CKD progression
  • Fluid overload at KRT initiation independently predicts death (PICARD, PICARD-era cohorts)

Koyner Ch 6 · Bouchard J et al. Kidney Int 2009;76:422 · Mehta RL et al. Kidney Int 2011 · Coca SG et al. Kidney Int 2009

AKI-D: the sickest subgroup you will follow for years

  • ~13% of ICU patients receive KRT; in the BEST Kidney cohort hospital mortality reached 60%
  • Modern trial cohorts: 90-day mortality ≈ 44% (STARRT-AKI, both arms)
  • Among survivors, 10–30% remain dialysis-dependent at hospital discharge
  • KRT dependence at 90 days is now a co-primary trial endpoint — and it was higher with accelerated initiation (10.4% vs 6.0%)
  • Most recovery happens in the first 1–4 weeks: wean deliberately (Lecture 9)

What to tell the family on day 1

"Dialysis here is usually temporary support while the kidney recovers, but it is also a marker of how sick the whole body is. We will know much more in two weeks."

Uchino S et al. JAMA 2005;294:813 · STARRT-AKI Investigators. N Engl J Med 2020;383:240 · Koyner Ch 6

Beyond discharge: post-AKI care is a clinical duty

What the survivor carries

  • Higher risk of CKD, ESKD, heart failure, hypertension and death
  • Risk is graded by stage, duration and number of episodes
  • Present even when creatinine returns to "normal" — the reserve is gone

What the discharge summary must contain

  • Peak stage, cause, whether KRT was used, discharge creatinine
  • Medication list reconciled: which nephrotoxins were stopped and whether to restart
  • A creatinine and urine-protein check within 3 months
  • Named follow-up: nephrology for stage 3, KRT-treated, or unrecovered patients

Koyner Ch 51 · See EJ et al. Kidney Int 2019;95:160 · KDIGO 2012 (Section 3.5)

06

Cases & special hosts

Contexts where the same criteria describe a completely different differential.

Koyner Ch 38–39
Case 1 · 68 M, septic shock from urinary source

ICU day 2. Norepinephrine 0.4 µg/kg/min, MAP 62. Receiving piperacillin-tazobactam and vancomycin. Abdomen distended.

Cr 1.1 → 2.4 mg/dL / 24 hUO 0.4 mL/kg/h × 8 hCVP 16Lactate 3.1Bladder pressure 19 mmHg

What KDIGO stage — and what are the four mechanisms operating simultaneously?

Think 45 seconds · answer on the next slide

Case 1 — four mechanisms, one creatinine

Stage 2

Cr 2.2× baseline. The UO criterion at 8 h does not yet qualify — reassess at 12 h.

Sepsis

Microcirculatory heterogeneity and inflammation — the dominant driver, and not fixed by more fluid.

Nephrotoxins

Check the vancomycin AUC; ask whether the combination is still needed once cultures return.

Congestion

IAP 19 mmHg with CVP 16 — renal perfusion pressure is MAP − IAP ≈ 43 mmHg. Decompress, do not fill.

Key point

Treat the whole picture: source control, restore perfusion pressure, stop what you can stop, and relieve the abdomen. No single intervention here is "the" treatment.

Koyner Ch 5, 8, 36 · Ronco Ch 45

Cardiac surgery-associated AKI

  • Incidence 20–30% by KDIGO; 1–2% need dialysis — but that 1–2% has mortality above 40%
  • Mechanisms: low output, haemolysis with free haemoglobin, systemic inflammation from bypass, atheroemboli, contrast from recent angiography, nephrotoxins
  • Preoperative risk scores (Cleveland Clinic / Thakar, Mehta) stratify before the incision
  • Bypass time and re-exploration are the strongest intraoperative predictors

What actually prevents it

Delay elective surgery ≥24–48 h after contrast, avoid intraoperative hypotension, restrictive transfusion, and apply the KDIGO bundle to biomarker-positive patients (PrevAKI). Not: "renal-dose" dopamine, mannitol, or fenoldopam.

Koyner Ch 39 · Thakar CV et al. J Am Soc Nephrol 2005;16:162 · Meersch M et al. Intensive Care Med 2017;43:1551 (PrevAKI)

Obstetric AKI

Pregnancy-specific causes

  • Preeclampsia / eclampsia / HELLP — the commonest
  • Haemorrhage: abruption, atony, previa; acute cortical necrosis in severe cases
  • Sepsis: chorioamnionitis, puerperal, septic abortion
  • TMA spectrum: aHUS (often postpartum), TTP (often antepartum)
  • Acute fatty liver of pregnancy

Management frame

  • Normal pregnancy creatinine is lower (≈0.4–0.8) — a "normal" 1.0 is already abnormal
  • Delivery is often the definitive therapy in preeclampsia/HELLP
  • Magnesium for seizure prophylaxis — dose-adjust in AKI, follow levels
  • Dialysis is safe in pregnancy; start earlier and dialyse longer (target BUN <45 mg/dL)
  • Do not forget obstruction — gravid uterus, stones

Koyner Ch 38 · Ronco Ch 40 · Fakhouri F et al. J Am Soc Nephrol 2017

07

Pitfalls & wrap-up

The two cognitive errors that account for most missed diagnoses on the consult service.

Koyner Ch 5

Two errors that cost kidneys

Pitfall 1 — premature closure on "ATN"

Glomerulonephritis, TMA, interstitial nephritis and obstruction all hide inside ICU AKI. Before writing "ATN", confirm you have looked at the sediment, checked platelets and LDH, reviewed every new drug, and imaged the bladder.

Pitfall 2 — trusting derived ratios in the wrong host

Liver disease and malnutrition lower BUN (a false "intrinsic" pattern); GI bleeding and steroids raise it (a false "prerenal" pattern). Sarcopenia flattens creatinine. Trimethoprim, cimetidine and cobicistat raise creatinine without touching GFR.

Koyner Ch 5 · Perazella MA. Clin J Am Soc Nephrol 2019

Quick poll

A 74 M, day 6 of ceftriaxone for pneumonia. Cr 1.0 → 3.1 mg/dL, urine output preserved at 1.1 mL/kg/h, bladder scan 30 mL, BP 145/85. Maculopapular rash on both legs. WBC 12 ×10⁹/L with 6% eosinophils; urine shows sterile pyuria and WBC casts.

A. Prerenal · B. Acute interstitial nephritis · C. Postrenal · D. Ischaemic ATN

Hands up — then justify your answer with one finding

Key takeaways

  • Stage with KDIGO using both creatinine and urine output — and write down which baseline you used.
  • Ask "will this reverse?" — transient versus persistent AKI predicts outcome better than the stage alone.
  • Expect ICU AKI to be multifactorial; septic AKI is inflammatory and microcirculatory, not simple ischaemia.
  • Sediment, urine chemistries and a bedside ultrasound belong in the first 24 hours — every time.
  • ~57% of ICU patients develop AKI and ~13% receive KRT; act with the urgency those numbers deserve.
  • AKI is a doorway to CKD: name the follow-up before the patient leaves your service.

References & further reading

Where to go deeper

  1. KDIGO AKI Work Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl. 2012;2:1–138.
  2. Hoste EAJ, Bagshaw SM, Bellomo R, et al. Epidemiology of acute kidney injury in critically ill patients: the multinational AKI-EPI study. Intensive Care Med. 2015;41:1411–1423.
  3. Uchino S, Kellum JA, Bellomo R, et al. Acute renal failure in critically ill patients (BEST Kidney). JAMA. 2005;294:813–818.
  4. Chawla LS, Bellomo R, Bihorac A, et al. Acute kidney disease and renal recovery: consensus report of the ADQI 16 Workgroup. Nat Rev Nephrol. 2017;13:241–257.
  5. See EJ, Jayasinghe K, Glassford N, et al. Long-term risk of adverse outcomes after AKI: a systematic review and meta-analysis. Kidney Int. 2019;95:160–172.
  6. Takasu O, Gaut JP, Watanabe E, et al. Mechanisms of cardiac and renal dysfunction in patients dying of sepsis. Am J Respir Crit Care Med. 2013;187:509–517.
  7. Koyner JL, Topf JM, Lerma EV, eds. Handbook of Critical Care Nephrology. Wolters Kluwer; 2021 (Ch 5–7, 38–40, 51).
Critical Care Nephrology · Two-Week Intensive

Thank you

Questions & discussion — bring a consult from this week and we will stage it together.

Next: Lecture 02 — Renal Hemodynamics, Shock & ICU Monitoring