The comprehensive evaluation of acute kidney injury requires systematic integration of clinical history, urinalysis interpretation, targeted laboratory studies, and appropriate imaging. This review synthesizes the diagnostic approach, emphasizing limitations of commonly used markers and the practical application of current quality metrics in clinical practice.
| Stage | Creatinine Change | Urine Output | Mortality Risk | Clinical Implications |
|---|---|---|---|---|
| Stage 1 | 1.5–1.9× baseline OR ≥0.3 mg/dL rise | <0.5 mL/kg/h for 6–12h | 2–5% | Early intervention window; reversibility likely |
| Stage 2 | 2–2.9× baseline | <0.5 mL/kg/h for ≥12h | 5–10% | Increased risk of complications |
| Stage 3 | ≥3× baseline OR ≥4 mg/dL rise OR RRT need | <0.3 mL/kg/h for ≥24h OR anuria ≥12h | 15–40% | High mortality; RRT often required |
KDIGO staging uses creatinine rise relative to baseline, but baseline is often unknown in hospitalized patients. Use lowest creatinine within the admission, recognizing this may underestimate AKI severity if baseline renal function was chronically low.
Creatinine is produced at steady state by muscle metabolism. AKI causes acute kidney dysfunction but creatinine may lag 24–48 hours even with sudden GFR loss. A patient with acute anuric obstruction may appear to have mild AKI on day 1 because creatinine has not yet risen.
In 30–40% of hospitalized AKI patients, true baseline creatinine is unavailable. A rise from 0.8 to 1.1 mg/dL is Stage 1; but if baseline was 2.0, the 1.1 may represent recovery from CKD.
The “creatinine paradox”: creatinine can fall in critically ill patients with declining GFR because muscle wasting exceeds tubular reabsorption decrease. Parallel cystatin C trending improves staging accuracy.
Massive creatinine rise (>4 mg/dL in 24 hours) may reflect muscle injury, not just kidney dysfunction. Urine dipstick positive for “blood” (actually myoglobin) without RBCs on microscopy. CK often markedly elevated (>5,000 IU/L). These patients are not truly in KDIGO Stage 3 AKI in the sense of intrinsic renal parenchymal damage; rather, they have severe myoglobinuria precipitating ATN.
The urine sediment is the single most useful diagnostic test in AKI. Obtain fresh urine (not catheter bag specimen) and examine within 30 minutes. Casts dissolve in alkaline urine and dilute specimens.
Brown, granular, “muddy” casts containing tubular epithelial cells, cellular debris, and hemosiderin pigment. During ATN, proximal tubule cells undergo necrosis and slough into the tubular lumen.
Specificity: 95% specific for ATN when present, but sensitivity only 40–60% (many ATN cases have no casts).
“Muddy brown” casts are pathognomonic for ATN when seen. Their absence does not rule out ATN. A perfectly normal UA (no cells, no casts) in a patient with AKI and preserved urine output is more consistent with prerenal AKI or early obstruction.
Casts containing neutrophils, sometimes with eosinophil predominance. Associations include drug-induced AIN (antibiotics, NSAIDs, PPIs, allopurinol), pyelonephritis with AKI, and less commonly lupus, sarcoidosis, or infection-related GN.
Eosinophiluria is present in only 20–30% of drug-induced AIN cases and can occur in other conditions (atheroemboli, parasitic infection). The standard UA dipstick does not detect eosinophils. Never use the absence of eosinophiluria to exclude drug-induced AIN. Ref: PubMed
Casts containing intact or fragmented RBCs, often with dysmorphic RBCs. >80% dysmorphic is highly specific for GN. Associated findings include proteinuria (>0.5–1 g/day), systemic signs (rash, arthralgias), and hypertension.
Casts and cells containing lipid droplets visible as doubly refractile under polarized light microscopy (maltese cross pattern). Almost always indicate nephrotic-range proteinuria (>3 g/day).
| Finding | Points |
|---|---|
| WBC casts | 3 |
| RBC casts | 2 |
| Renal tubular epithelial cells (≥1/hpf) | 1 |
| Granular casts (coarse or heavy) | 1 |
Interpretation: Score ≥2: 90% sensitivity, 83% specificity for AIN. Score 0–1: Rules out AIN with high negative predictive value.
Ref: Perazella MA, Coca SG. Clin J Am Soc Nephrol. 2012;7(1):167–174. PubMed
Fractional excretion indices measure the fraction of filtered sodium (or urea, uric acid) that appears in the urine. They reflect tubular reabsorptive function, not the etiology of AKI.
Formula: FeNa = (Urine Na × Plasma Cr) / (Urine Cr × Plasma Na) × 100
| Confounder | Mechanism | Example |
|---|---|---|
| Loop/thiazide diuretics | Block NaCl reabsorption → high FeNa | Patient on furosemide with sepsis-induced ATN: FeNa 5% |
| CKD baseline | Chronic tubular dysfunction | CKD Stage 4 + contrast AKI: FeNa 0.5% despite ATN |
| Bicarbonaturia | HCO3 excretion couples with Na+ | Metabolic alkalosis + volume depletion: FeNa 2–3% in prerenal |
| Sepsis | Endotoxin-mediated tubular dysfunction | Septic shock with ATN: FeNa <1% |
| Contrast nephropathy | Direct tubular toxicity | Contrast AKI: FeNa 1–3% |
| Glomerulonephritis | Primary glomerular lesion activates RAAS | Post-infectious GN: FeNa <1% |
| Rhabdomyolysis | Pigment toxicity + intact tubular response | Crush injury: FeNa 0.5% |
FeNa sensitivity for prerenal AKI is only 80–90%, and specificity for ruling out ATN is 60–70%. Do not use FeNa as your sole basis for diagnosis.
Formula: FeUrea = (Urine Urea × Plasma Cr) / (Urine Cr × Plasma Urea) × 100
Theoretical advantage: urea reabsorption is less affected by diuretics. However, clinical validation is mixed and it is not routinely recommended.
Preliminary data in small cohorts suggest FeUA may identify uricosuric AKI, but no large prospective trials exist. Not recommended for routine clinical use.
To distinguish prerenal from post-renal AKI, rely on imaging (renal ultrasound) and clinical context, not FeNa. FeNa is useful only for the prerenal-vs-ATN decision and should always be interpreted with caution in diuretic-treated patients.
Renal ultrasound is the first-line imaging test for suspected AKI when obstruction is in the differential. Key assessments:
Negative ultrasound does not exclude obstruction. If high clinical suspicion persists (e.g., history of stone, flank pain), consider CT or direct visualization (cystoscopy/ureteral catheter).
Formula: RI = (Peak systolic velocity − End-diastolic velocity) / Peak systolic velocity
Limitation: RI is nonspecific. It does not differentiate etiologies. RI <0.8 at AKI onset predicted 80% recovery; RI >0.8 predicted only 40% recovery (not prospectively validated).
| Component | Definition |
|---|---|
| Numerator | Patients who develop KDIGO Stage 2+ AKI during hospitalization (≥2× baseline OR ≥4 mg/dL rise) |
| Denominator | All patients age ≥18 hospitalized for ≥2 calendar days |
| Exclusions | ESRD on dialysis; admission SCr >2.5 mg/dL; kidney transplant this admission; palliative care |
| Rate | Stage 2+ AKI cases per 1,000 hospital days |
Age, sex, eGFR at admission, diabetes status, CHF history, major surgery, sepsis flag, length of stay, and ICU admission. Observed-to-Expected Ratio (O:E) >1.0 means the hospital has more AKI than expected for its risk profile.
| Test | Indication | Timing | If Positive |
|---|---|---|---|
| UA with sediment | All AKI | Within 24h | Directs further workup |
| FeNa | Prerenal vs. ATN (no diuretics) | 24–48h | <1% prerenal; >2% ATN |
| BUN/Cr ratio | Prerenal assessment | 24h | >20 supports prerenal |
| Renal US | Obstruction suspected | Urgent (24–48h) | Hydronephrosis → urology |
| CT abdomen | US negative, high suspicion | 48–72h | Confirms obstruction |
| Serologies | RBC casts or hematuria + AKI | 48–72h | GN → nephrology referral |
| Renal biopsy | Unexplained AKI | 72h+ | Diagnoses GN, sarcoidosis, TMA |
Cardiorenal AKI: Combine UA findings with clinical HF signs. Check BNP/NT-proBNP. Consider euvolemic ultrafiltration if diuretic-resistant congestion.
Hepatorenal Syndrome: AKI in cirrhotic patient with portal hypertension. FeNa <0.1% (very avid Na+ reabsorption). Oliguria, low urine sodium (<10 mEq/L). Normal kidney size on US. Treatment: vasoconstrictors (terlipressin, norepinephrine) + albumin, NOT diuretics or fluids.