Urine Studies for Diagnosing Acute Kidney Injury (AKI)

Urine studies offer a way to detect acute kidney injury earlier and more specifically than the standard blood test for creatinine, which often rises only after significant damage has already occurred. A growing toolkit of urine-based measurements now ranges from old-fashioned microscopy and electrolyte ratios to newer protein biomarkers and even a functional stress test using a common diuretic. None of these tests has fully replaced serum creatinine in everyday practice, but each fills a different diagnostic gap, and the field is evolving fast enough that what clinicians rely on today looks quite different from a decade ago.

What Traditional Urine Microscopy Still Reveals

Before any fancy biomarker assay existed, clinicians spun a tube of urine in a centrifuge and looked at the sediment under a microscope. That technique remains useful. A systematic review found that the presence and number of renal tubular epithelial cells, tubular epithelial cell casts, and granular casts in the urine sediment help diagnose acute tubular necrosis and may predict more severe kidney damage, including failure to recover kidney function and the need for dialysis.¹ A “bland” sediment, meaning one without these cellular elements, points more toward a prerenal cause, where the kidneys themselves are structurally intact but not getting enough blood flow. The limitation is obvious: reading urine sediment well requires skill and practice, and interpretations can vary between operators. Still, it costs almost nothing and can be done at the bedside, which keeps it relevant even in the biomarker era.

Fractional Excretion of Sodium and Urea

For decades, clinicians have measured how much sodium or urea the kidneys excrete to distinguish prerenal AKI (a blood flow problem) from intrinsic AKI (actual kidney tissue damage). The fractional excretion of sodium (FENa) compares how much sodium your kidneys filter to how much they let pass into the urine. A very low value suggests the kidneys are hanging on to sodium tightly, which is what healthy tubules do when blood flow drops. A high value suggests the tubules are damaged and can no longer reabsorb sodium properly.

The fractional excretion of urea (FEUrea) was introduced as an alternative, mainly because diuretics like furosemide throw off sodium handling and make FENa unreliable. The hope was that urea excretion would remain interpretable even when patients are on diuretics. A 2024 meta-analysis pooling data from over 900 patients found that FEUrea at a cutoff above 35% had a pooled sensitivity of about 66% and specificity of about 75% for distinguishing intrinsic from prerenal AKI. In patients not on diuretics, there was no meaningful difference in accuracy between FEUrea and FENa. In patients receiving diuretics, FEUrea was less sensitive than FENa but more specific. The authors concluded that FEUrea has limited overall utility and does not provide a clearly superior alternative to FENa, even in the diuretic population it was designed for.²

An earlier study reached a compatible but sharper conclusion: in patients without diuretics, FENa was the better performer, and in patients on diuretics, neither test could make the distinction accurately enough to be relied upon alone.³ The upshot is that these classic urine electrolyte tests remain part of the initial workup but are best treated as one data point among many rather than a definitive answer.

Why Urine Eosinophils Are Less Useful Than Textbooks Suggest

If you look in older clinical references, you will find a recommendation to check for eosinophils in the urine when acute interstitial nephritis (AIN) is suspected, typically as a drug-induced allergic reaction in the kidneys. The idea was that a urine sample with more than 1% eosinophils pointed toward AIN. A study that tested this directly found the reality is disappointing: at that 1% cutoff, the sensitivity was only about 31% and specificity about 68%. The positive likelihood ratio was essentially 1.0, meaning the test did not shift the probability of AIN in either direction. Even at a higher 5% cutoff, urine eosinophils performed poorly in separating AIN from acute tubular necrosis or other kidney diseases.⁴ This is one of those clinical tests that persists by tradition more than by evidence. A kidney biopsy remains the most reliable way to confirm AIN when the clinical picture is unclear.

NGAL as an Early Warning Signal

Neutrophil gelatinase-associated lipocalin, known as NGAL, is one of the most studied urine biomarkers for AKI. It is a small protein that gets released by injured kidney tubule cells, and it rises in the urine within hours of an insult, well before serum creatinine begins to climb. Researchers have described it as a potential “troponin-like” early marker for AKI, drawing an analogy to the blood test cardiologists use to detect heart attacks quickly.⁵ NGAL has been studied across a range of clinical scenarios including cardiac surgery, kidney transplantation, contrast-related kidney injury, and intensive care.⁶

A recent prospective study using a point-of-care NGAL test in a South Asian hospital found that urinary NGAL above 125 ng/mL was highly specific (about 90%) and moderately sensitive (about 82%) for identifying patients with a sustained rise in creatinine. Elevated NGAL was also associated with roughly a fourfold increase in the odds of death.⁷ However, NGAL is not perfect. Its levels can be elevated by urinary tract infections (because neutrophils produce it too), and it can be influenced by pre-existing chronic kidney disease. In kidney transplant recipients, for instance, urinary NGAL was related to signs of tubular injury on biopsy, but the sensitivity was too low to be a reliable standalone monitoring tool in that population.⁸

KIM-1 and Tubular Damage Detection

Kidney Injury Molecule-1 (KIM-1) is a protein that sits on the surface of proximal tubule cells and is barely detectable in healthy kidneys. After ischemic injury, its expression skyrockets. In one of the early human studies, urinary KIM-1 levels were significantly higher in patients with ischemic acute tubular necrosis than in patients with other forms of acute kidney failure or chronic kidney disease. A one-unit increase in normalized KIM-1 was associated with roughly a 12-fold increase in the odds of having acute tubular necrosis.⁹ This makes KIM-1 particularly good at fingerprinting one specific type of AKI. Its main drawback is speed: KIM-1 tends to rise more slowly than NGAL, peaking later after the initial insult, which limits its usefulness as the earliest possible alarm.

In cardiac surgery patients, urinary KIM-1 rose progressively throughout the operation and into the ICU stay, and patients who went on to develop AKI had significantly higher KIM-1 levels at multiple time points compared with those who did not.¹⁰ By contrast, another marker tested in that study, interleukin-18, showed similar rises in both groups until relatively late, making it less useful for early detection in that context.

The FDA-Approved Cell Cycle Arrest Test

In 2014, the US Food and Drug Administration cleared a urine test that combines two proteins involved in cell cycle arrest: tissue inhibitor of metalloproteinase 2 (TIMP-2) and insulin-like growth factor binding protein 7 (IGFBP7). The test multiplies their concentrations together and is marketed under the brand name NephroCheck. It was approved specifically to help identify critically ill patients at risk of developing moderate to severe AKI within 12 hours.¹¹

An interesting wrinkle about these markers is that their elevation in the urine during AKI does not come from the kidneys ramping up production of the proteins in response to stress, as originally assumed. Animal research showed that the increase comes instead from a combination of more protein being filtered through damaged glomeruli, less being reabsorbed by injured tubules, and some leaking out of damaged proximal tubule cells.¹² This matters because it means the test is picking up structural damage rather than a targeted stress response, which has implications for how the result should be interpreted.

The clinical outcome data are encouraging. A consensus statement noted that a TIMP-2 × IGFBP7 value above a commonly used threshold was associated with about double the hazard of death or the need for kidney replacement therapy among patients with AKI.¹³

The Furosemide Stress Test

Rather than measuring a molecule in the urine, the furosemide stress test (FST) measures what the kidneys do when challenged with a single standardized dose of furosemide, a powerful loop diuretic. The logic is simple: furosemide must be actively secreted by proximal tubule cells to reach its site of action. If the tubules are severely damaged, the drug cannot get where it needs to go, and urine output stays low. In the original development study, a urine output below about 200 mL over the two hours following the dose predicted progression to severe AKI with roughly 87% sensitivity and 84% specificity.¹⁴

A meta-analysis confirmed these findings across multiple studies, reporting pooled sensitivity and specificity of about 81% and 88% for predicting AKI progression, and about 84% and 77% for predicting the need for dialysis. The test performed particularly well in patients with early-stage AKI and was not thrown off by underlying chronic kidney disease.¹⁵ The FST has the appeal of being cheap, widely available, and conceptually transparent: you give a drug that requires working tubules, and if the tubules respond, the prognosis is better. The obvious caveat is that it requires the patient to be adequately hydrated, and volume status must be carefully managed around the test.

Subclinical AKI and the Limits of Creatinine

One of the more consequential ideas to come out of the biomarker research is subclinical AKI: kidney damage that is real and detectable by urine biomarkers but has not yet caused a measurable rise in serum creatinine or a drop in urine output. Current clinical definitions of AKI from the KDIGO guidelines rely entirely on creatinine and urine output, which means they miss this early damage window.¹⁶ Tubule damage markers can flag this injury before filtration function is lost, potentially allowing earlier intervention.¹⁷ The challenge is that, so far, there is no consensus on how to treat subclinical AKI differently from “no AKI.” Identifying damage earlier only helps if you can actually do something about it, and the interventional trials are still catching up to the diagnostic science.

How Context Changes Performance

The accuracy of urine biomarkers varies considerably depending on the clinical scenario. Cardiac surgery is the most intensively studied setting, partly because the timing of the insult is known precisely (the moment of cardiopulmonary bypass). In cardiac surgery patients, combining KIM-1, NAG, and NGAL at three hours after the operation improved diagnostic accuracy over any single marker, with an area under the curve reaching about 0.78 compared with roughly 0.59 to 0.68 for the individual biomarkers alone.¹⁸ Another cardiac surgery study found that L-FABP, a fatty acid-binding protein released from injured tubules, rose in the AKI group as early as the immediate post-operative period.¹⁹

In sepsis, the picture is more complicated. Sepsis-associated AKI involves a tangled mix of inflammation, microvascular dysfunction, and direct tubular damage that makes single biomarkers harder to interpret. A study of emergency department patients with sepsis found that both NGAL and TIMP-2 × IGFBP7 were significantly higher in those who developed AKI than in those who did not, but none of the tested biomarkers could distinguish survivors from non-survivors.²⁰ The heterogeneity of sepsis, differences in the infecting organism, the patient’s immune response, the timing of treatment, creates noise that limits how well any single urine marker can perform.²¹

Contrast-induced AKI after procedures involving iodinated dye (like coronary angiography) is another scenario where early biomarker detection could theoretically change management. One study found that urinary interleukin-18 and NGAL rose significantly at 24 hours in patients who developed contrast nephropathy, about a day before serum creatinine flagged the problem.²² That time gap matters because it opens a theoretical window for protective measures such as aggressive hydration, though the evidence for interventions once contrast has already been given is limited.

Pediatric Complications

Children are not small adults when it comes to urine biomarkers. Normal baseline levels of NGAL, KIM-1, and other markers change with age and differ between boys and girls, which means adult reference ranges cannot be applied to children. A study of healthy children aged 3 to 18 found that NGAL was significantly higher in girls than in boys across all age groups, and that KIM-1 levels increased with age, peaking in adolescence.²³ A more recent effort to build reference percentile curves for seven urinary biomarkers of kidney injury confirmed that all were age-dependent, and that NGAL and several others also varied by sex.²⁴ Without pediatric-specific thresholds, there is a real risk of misinterpreting a normal value as abnormal or vice versa.

What Can Interfere with Results

Pre-analytical variables can muddy the picture. A study of cardiac surgery patients found that the presence of leukocyte esterase and nitrites on a standard urine dipstick (markers of urinary tract infection) was associated with higher NGAL readings in samples collected before surgery even began, which means an unrecognized UTI could falsely elevate NGAL and mimic AKI. Hematuria raised levels of all four biomarkers tested, particularly in post-operative samples. Dipstick proteinuria correlated with all four biomarkers and, when accounted for statistically, reduced their diagnostic accuracy.²⁵ The practical lesson is that a simple dipstick urinalysis should accompany any biomarker measurement so that confounders like infection, blood, and protein can be identified and factored into the interpretation.

Combining Biomarkers into Panels

Because no single biomarker captures every dimension of AKI, researchers have explored whether combining several into a panel improves accuracy. In one cardiac surgery study, certain two-marker combinations, pairing a tubular injury marker with a filtration marker, reached areas under the curve of around 0.83 to 0.86, which is respectably high. However, when the investigators tried to use these panels to actually classify individual patients into high-risk and low-risk groups, the combinations did not significantly outperform the best single biomarker.²⁶ This is a recurring theme: multi-biomarker panels improve the curve on statistical plots but do not always translate into better decisions at the bedside.

The Cost-Effectiveness Question

A biomarker that works is only half the battle. It also has to be worth paying for. The economic evidence is mixed and, frankly, immature. A UK health technology assessment found that cost-effectiveness results for both NephroCheck and NGAL tests were “highly uncertain” and “largely speculative,” with some modeling scenarios showing the tests saved money and others showing they cost more while delivering fewer quality-adjusted life years than standard care.²⁷ Under base-case assumptions in another UK-based analysis, the NephroCheck test came in at over £100,000 per quality-adjusted life year, and various NGAL assays were even higher, all well above the standard willingness-to-pay threshold.²⁸

Pediatric cardiac surgery may be one area where the math works out better. A cost-effectiveness analysis in that population found that a urinary L-FABP strategy was cost-effective compared with the standard creatinine-only approach, adding about 1.4 quality-adjusted life years at an incremental cost of roughly $6,000 per QALY gained.²⁹ The difference in economic outcomes likely reflects the higher stakes in children, where early detection can prevent long-term damage that carries costs for decades. For adult ICU populations, the field is still waiting for better trial data before the economics become clear.

Urinary Metabolomics and Where the Science Is Headed

Beyond measuring one or a handful of proteins, researchers have begun analyzing the entire landscape of small molecules in the urine using untargeted metabolomics. A recent study of cardiac surgery patients identified 45 metabolites that were altered in those who developed AKI, with a polyamine breakdown product called N-acetylputrescine standing out as a strong discriminator. Several other polyamine catabolites were also elevated, suggesting that a specific metabolic pathway, polyamine breakdown, ramps up during kidney injury. Some of these metabolites correlated with AKI severity.³⁰ Meanwhile, animal research using detailed time-course sampling has identified completely novel candidate biomarkers such as PARK7 and CDH16, proteins whose concentrations change continuously during sepsis-associated AKI and track with creatinine levels when validated in human patients.³¹ These approaches are still firmly in the research phase, far from the bedside, but they represent a shift from testing individual candidate molecules one at a time to scanning the urine for entire patterns of injury. If the technology matures, future tests might detect not just whether AKI is present but what kind of injury is occurring and how it is likely to evolve.