The timing of investigations depends on the clinical condition of the patient and the rate and magnitude of change in serum creatinine. In asymptomatic or mildly symptomatic patients, elevations in serum creatinine can be systematically evaluated. However, in severely ill patients with multi-organ involvement, an early renal biopsy should be considered to expedite diagnosis and treatment. If there is evidence of fluid overload or uraemic signs and symptoms, treatment should be instigated in parallel with urgent investigations.
Current serum creatinine values should be compared with previous serum creatinine values, if available, to distinguish acute from chronic changes. If previous values are not available, all values outside of the reference range should be treated as acute changes. The Acute Kidney Injury Network criteria should be checked to determine whether patients meet the criteria for acute kidney injury (AKI):
Absolute increase in serum creatinine of ≥26.4 micromol/L (≥0.3 mg/dL), or
Percentage increase in serum creatinine of ≥50% (1.5-fold from baseline), or
Reduction in urine output (documented oliguria of <0.5 mL/kg per hour for >6 hours).
If the elevation in serum creatinine is acute, the magnitude and rate of elevation in serum creatinine should be determined, as this provides clues as to the underlying cause, and guides appropriate management:
The typical rise in serum creatinine in AKI is 88 to 177 micromol/L/day (1 to 2 mg/dL/day).
Serum creatinine rises >88 to 177 micromol/L/day (>1 to 2 mg/dL/day) should prompt suspicion of rapidly progressive glomerulonephritis or rhabdomyolysis. Rises as high as 442 micromol/L/day (5 mg/dL/day) may be seen in rhabdomyolysis.
An acute increase in serum creatinine against the background of chronic kidney disease (CKD) may indicate progression of the underlying condition. However, other causes of elevated serum creatinine should be excluded first. Fluid balance issues, medications, radiocontrast exposure, and worsening of comorbid conditions are the most common causes and may be reversible.
Most of the underlying causes produce few or no specific symptoms. Uraemia and an altered mental status may occur and are more commonly seen in advanced AKI or in advanced CKD.
Symptoms of uraemia are often vague. They include nausea, vomiting, fatigue, reduced appetite, muscle cramps, pruritus, and altered mental status.
A history of trauma, severe vomiting, or diarrhoea should prompt suspicion of volume depletion.
Joint pain, rash, and haemoptysis should prompt suspicion of vasculitis.
Fever and sore throat should prompt suspicion of streptococcal infection (causing glomerulonephritis).
Jaundice should prompt suspicion of hepatitis B or C infection, or hepatorenal syndrome.
Shortness of breath with peripheral oedema may indicate heart failure.
Weight loss should prompt suspicion of an underlying malignancy.
Obstructive (voiding) and irritative (storage) symptoms should prompt suspicion of obstructive uropathy.
Severe loin pain with frank haematuria should prompt suspicion of acute renal infarction or nephrolithiasis.
A history of polyuria, polydipsia, polyphagia, weakness, and/or weight loss should prompt suspicion of diabetic ketoacidosis.
Pregnancy should prompt suspicion of pre-eclampsia or exacerbation of underlying kidney disease.
Past medical history
Chronic renal failure.
Previous acute renal failure.
Other underlying conditions that may cause raised serum creatinine include hypertension (hypertensive nephropathy), diabetes (diabetic nephropathy), autoimmune diseases (vasculitis), cirrhosis (hepatorenal syndrome), lymphoproliferative disorders (cryoglobulinaemia), paraproteinaemias, and infections (post-infectious glomerular nephritis, cryoglobulinaemia).
Recent surgery may indicate hypovolaemia, renal ischaemia (due to clamping of arteries during cardiac surgery), or renal transplant rejection. Multiple cholesterol emboli syndrome can occur following arterial manipulation, vascular surgery, stent placement, or cardiac catheterisation. A small rise in serum creatinine following kidney donation or unilateral or partial nephrectomy is expected and should resolve.
A thorough drug history is important, as the range of drugs known to cause elevated serum creatinine is extremely wide.
Cimetidine, gentamicin, fibric acid derivatives other than gemfibrozil, and trimethoprim inhibit creatinine secretion.
ACE inhibitors and angiotensin receptor blockers increase serum creatinine levels by 20% to 30% due to pre-renal effects.
Penicillamine, sodium aurothiomalate, non-steroidal anti-inflammatory drugs (NSAIDs), captopril, mitomycin C, and ciclosporin can cause glomerulonephritisas, as can heroin.
Many medications are known to cause acute interstitial nephritis.
Aminoglycosides, amphotericin-B, chemotherapeutic agents (e.g., cisplatin), NSAIDs, and radiocontrast media can cause acute tubular necrosis.
Some patients have an increase in serum creatinine within a few months after starting fenofibrate, an entity known as fenofibrate-associated creatinine increase.
Creatine is often taken as a supplement to boost muscle mass and increase athletic performance. Prolonged intake of creatine supplementation of >10 g/day may increase serum creatinine concentrations, but is unlikely to affect estimates of creatinine clearance. The serum creatinine levels should return to baseline values upon discontinuing supplemental creatine.
A vegetarian diet is associated with decreased generation of creatinine, and ingestion of cooked meat causes a transient increase in serum creatinine.
The examination may be unremarkable in many patients, but is most useful to differentiate systemically unwell patients (who require immediate treatment and early renal biopsy) from other patients who can be investigated systematically.
Patients with shock appear unwell, with decreased blood pressure, increase in heart rate, and increased respiratory rate with a possible decrease in oxygen saturations or reduced level of consciousness.
Uraemic signs such as asterixis or altered mental status, and/or signs of fluid overload (e.g., pulmonary or peripheral oedema), indicate severe renal failure requiring immediate treatment. Early renal biopsy to establish the cause should be considered in these patients.
The presence of rash should prompt suspicion of vasculitis or a microangiopathy. Ascites should prompt suspicion of cirrhosis. Jaundice should prompt suspicion of cirrhosis or hepatitis B or C infection. Peripheral oedema should prompt suspicion of heart failure.
The appearance of cutaneous lesions, 'thrash toes, blue toes', pancreatitis, stroke, ischaemic bowel, or angina should prompt suspicion of multiple cholesterol emboli syndrome.
Initial investigations include:
Urea/serum creatinine ratio can help distinguish pre-renal from renal causes. A ratio >100 (for SI units; >20 for conventional units) suggests pre-renal causes, whereas a ratio <40 (for SI units; <10 for conventional units) suggests renal causes.
Urinalysis provides useful diagnostic clues:
Isolated proteinuria suggests nephrotic syndrome, diabetic nephropathy, or pre-eclampsia
Isolated haematuria suggests nephrolithiasis
Proteinuria and haematuria without other abnormalities suggests acute interstitial nephritis.
Further urine tests include microscopy, specific gravity, osmolality, random urine sodium, and fractional excretion of sodium.
Normal or hyaline casts, specific gravity >1.020, osmolality >500 mOsm/kg H₂O, random urine sodium (UNa) <20 mEq/L, and a fractional excretion of sodium (FENa) <1% suggest pre-renal pathology. FENa may be calculated using the following formula: (urine sodium x plasma creatinine)/(plasma sodium x urine creatinine) x 100.[ Fractional Excretion of Sodium (SI units) ]
Haematuria, proteinuria, red blood cell casts, epithelial cell casts, waxy casts, granular casts, UNa >20 mEq/L, and FENa <1% suggest glomerulonephritis.
Muddy brown granular casts, epithelial cell casts, specific gravity approximately 1.010, UNa >20 mEq/L, and FENa >1% suggest acute tubular necrosis.
Myoglobin casts suggest rhabdomyolysis.
Eosinophiluria suggests atheroembolic disease (cholesterol emboli).
Crystals suggest a crystal-induced nephropathy. Examples include crystalluria in tumour lysis syndrome, calcium oxalate (ethylene glycol intoxication), medications (aciclovir, indinavir, sulfadiazine).
Other investigations are ordered depending on the clinical features and include the following.
Serum creatine kinase: elevated in rhabdomyolysis and multiple cholesterol emboli syndrome. Mild elevation is seen in patients taking creatine supplements.
Blood, urine, and sputum cultures if sepsis is suspected.
Renal biopsy: should be considered as an initial test in patients with uraemic symptoms or multi-organ involvement; it provides a definitive diagnosis of underlying renal causes. In all other patients, renal biopsy is reserved for cases in which confirmation of the diagnosis is required after a systematic investigation is completed.
Serology for underlying infections, including HIV, hepatitis B and C.
Vasculitis markers: antiglomerular basement membrane antibodies (positive in Goodpasture syndrome), antineutrophil cytoplasmic antibodies (positive in small-vessel vasculitis, e.g., granulomatosis with polyangiitis [formerly known as Wegener's granulomatosis], polyarteritis nodosa), and antinuclear antibodies or anti-double-stranded DNA antibodies (positive in systemic lupus erythematosus).
Antistreptolysin O or antibodies to streptokinase or hyaluronidase should be ordered if post-streptococcal glomerulonephritis is suspected.
Complement titres are low in post-infectious glomerulonephritis, systemic lupus erythematosus, subacute bacterial endocarditis, cholesterol embolisation, essential mixed cryoglobulinaemia, and membranoproliferative glomerulonephritis.
Protein electrophoresis should be ordered if paraproteinaemia is suspected and if it shows a protein spike corresponding to the monoclonal protein.
Uric acid is elevated in tumour lysis syndrome or pre-eclampsia.
HbA1c is useful to assess glycaemic control in people with diabetes.
The American College of Radiology provides guidelines for the selection of appropriate imaging modalities.
Renal Doppler ultrasound should be performed if renal artery stenosis is suspected.
Renal angiogram should be performed in cases of suspected renal artery thrombosis.
Computed tomography and a magnetic resonance venogram should be performed in cases of suspected renal vein thrombosis or traumatic renal infarction.
Renal and urinary tract ultrasound should be performed if obstructive uropathy is suspected.
Estimation and measurement of glomerular filtration rate (GFR) and creatinine clearance
Estimated GFR (eGFR) and creatinine clearance using serum creatinine
The most accurate method for calculating GFR is by measuring the clearance of exogenous filtration markers, such as iothalamate, iohexol, or inulin. However, this is expensive and requires exposure to radiation and compliance with strict regulatory guidelines. In practice, therefore, creatinine clearance is used to estimate GFR. Creatinine is freely filtered, has minimal tubular secretion and absorption, is simple and inexpensive to measure from random blood samples, and has relatively good accuracy. A rise in serum creatinine is used as a marker of reduced GFR. It varies inversely with GFR, but the relationship is not linear.
The use of serum creatinine as an indirect filtration marker is limited by the following factors:
Bias and non-specificity affecting creatinine measurement
Alterations in circulating serum creatinine produced by non-renal disease states
Differences in GFR range and creatinine production in healthy people compared with people with CKD.
As a result of these confounding factors, there is a risk of overestimating the GFR, and the magnitude of the overestimation is not predictable.
Equations for estimating GFR
Equations for estimating GFR from serum creatinine levels are mainly used for staging CKD and should not be used to interpret acute increases in serum creatinine. Correction factors for black people have generally been derived from studies in African-American people.
Available equations include:
Cockcroft-Gault formula[ Creatinine Clearance Estimate by Cockcroft-Gault Equation ]
Four-variable Modification of Diet in Renal Disease (MDRD) formula[ Glomerular Filtration Rate Estimate by the Abbreviated MDRD Study Equation (SI units) ]
Isotope dilution mass spectrometry (IDMS)-traceable 4-variable MDRD equation (MDRD-IDMS)[ Glomerular Filtration Rate Estimate by the IDMS-Traceable MDRD Study Equation ]
Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation[ Glomerular Filtration Rate Estimate by CKD-EPI Equation ]
Mayo Clinic equation
Lund-Malmo revised (LMR)
Full age spectrum (FAS) (can be used in children and adults of all ages)
Berlin initiative study 1 and 2 (BIS1, BIS2).
A comparison of GFR estimation equations reveals that the MDRD formula performs well in populations with a low range of GFR, and often outperforms the Cockcroft-Gault equation. Both equations have lower precision in high-GFR populations, and the MDRD equation underestimated the GFR in some studies. In addition, the Cockcroft-Gault, simplified MDRD, and CKD-EPI equations showed high variability in a study of critically ill patients, and are not recommended in the intensive care unit setting.
An evaluation of four different eGFR equations (CKD-EPI, LMR, FAS, BIS1) found that none was diagnostically superior in adults over 65 years of age with varying degrees of kidney impairment.
Pitfalls in the use of eGFR values
eGFR overestimates GFR in low GFR states (due to increased tubular secretion of creatinine).
eGFR is inaccurate in high GFR states (due to the lack of actual eGFR values above 60 mL/minute/1.73 m²).
Studies suggest that the Cockroft-Gault and MDRD formulas correctly assigned only 64% and 62% of patients, respectively, to their actual CKD classification GFR group. Based on US National Health and Nutrition Examination Surveys (NHANES 1988-1994; 1999-2004) and US population census data (2000), this suggests that around 10 million people (38%) may have been misclassified in the US.
Serum creatinine and eGFR may not be equivalent in all clinical situations. In contrast media-induced nephropathy, an increase in serum creatinine, but not eGFR, was predictive for long-term mortality, with a threshold of 44.2 micromol/L (0.5 mg/dL) or more indicating worse prognosis.
The controversy regarding the optimal method to estimate GFR for disease detection and monitoring is unresolved. Comparisons of GFR estimation, using the CKD-EPI equation and other creatinine-based equations in different populations found that these equations are not applicable to all populations and need to be individually validated prior to their routine use.
Prediction of renal function is often inaccurate in individuals with class 2/3 obesity (BMI ≥35 kg/m²). Body cell mass GFR (BCM GFR), calculated from total body electrical impedance analysis that utilises an estimate of fat-free body mass, predicts creatinine clearance more accurately than traditional formulas in this patient population.
Use of this content is subject to our disclaimer