Dr Graham Jones
Department of Chemical Pathology

Creatinine

Physiology

Creatinine is spontaneously formed from creatine in muscle and released into the bloodstream at a fairly constant rate. It is removed from the body by largely by glomerular filtration although additionally there is a small amount of creatinine removed by renal tubular secretion and gastro-intestinal losses. By far the major determinants of serum creatinine concentration are muscle mass and glomerular filtration rate (GFR). As muscle mass is generally constant within an individual over short time periods, changes in serum creatinine are useful markers of  changes in GFR.
In any individual with stable renal function the scatter of serum creatinine concentrations is very small and can be described with a Coefficient of Variation (CV) of about 4%. By contrast there is a wide variation in creatinine concentrations between individuals. This means that population-based reference intervals are often a poor tool to identify renal impairment with comparison to a patient's own previous serum creatinine measurement providing a much more sensitive tool for detection of changes in renal function. In general changes in an individual's serum creatinine of 15% or more are highly likely to indicate a significant change in renal function.

Pathology

Decreases in GFR are marked by a rise in serum creatinine. The relationship is described mathematically as a rectangular hyperbola, meaning that as the GFR falls by half, there is a doubling in the serum creatinine. This means that by the time a rise in serum creatinine is detected, there may be 30 - 50% fall in GFR. Thus serum creatinine is a very specific marker of falling GFR (ie there are no other causes of a significant rise in serum creatinine) but it is not very sensitive for small decreases in GFR.
With increasing adult age there is generally a significant fall in GFR which is nearly matched by a concomitant fall in muscle mass. The combined effect of these changes is that serum creatinine changes very little with advancing age into the geriatric range, in spite of decreasing GFR. Estimation of the GFR with a formula such as the Cockroft and Gault equation often allows better detection of decreased GFR than interpretation of serum creatinine alone.

Urine Creatinine

Creatinine in urine may be measured for a number of purposes. In a 24 hour urine sample urine creatinine can be used as a rough marker of completness of the sample. Samples with a low 24 hour urine creatinine (and low urine volume) may be due to incomplete collection and samples with high 24 hour urine creatinine may be due to collection for longer than a single day. Note that these findings may be due to low or high muscle mass for the patient respectively. Urine creatinine may also be measured in a 24 hour urine sample to calculate the creatinine clearance, although this method of estimating creatinine clearance is considered less accurate than calculations such as the Cockroft and Gault equation in most circumstances.
Creatinine is often measured in spot urine samples as a marker of the effect of hydration on other analytes, for example measurement of the albumin:creatinine ratio. A low spot urine creatinine generally indicates high urine output, for example after high fluid intake and high urine creatinine may be caused by the body's response to dehydration. In the setting of urine drug testing a very low urine creatinine may be due to dilution of the sample with water or other fluids.

Serum Creatinine Assay Changes

On Monday 13th September 2004 the SydPath serum creatinine assay was changed in a number of ways as described below (full details in attached word document).

REPORTING UNITS: The reporting units are changing from mmol/L (millimoles per litre) to umol/L (micromoles per litre). This increases the numerical values of the results by a factor of 1000 and brings SydPath in line with international standards.

STANDARDISATION: The Sydney serum creatinine assay will be re-standardised in line with International Reference Methods and other Australian laboratories. The results will change according to the following equation:

Serum Creatinine (new, umol/L) = Serum Creatinine (old, mmol/L) x 1086 – 26  umol/L

This will create a slight fall in normal and low results (by 10 to 20 umol/L) and a slight rise in higher results (results over 400 umol/L increasing by 10 – 30 umol/L).  
(use the Result Converter to convert from "Old" to "New" results) or ("New" to "Old" results)

Moree and Narrabri laboratories: the restandardisation will produce results which are 10 umol/L (0.01 mmol/L) lower than current results at all creatinine concentrations.

REFERENCE INTERVAL ADJUSTMENT:
New Adult Reference Intervals are introduced reflecting the change is assay standardisation.
Male: 60 – 120 umol/L
Female: 40 – 90 umol/L

Other Notes:

Serum Creatinine Reporting: The serum creatinine will be reported to the nearest umol/L. This does not imply improved precision in the results. When within-patient patient variation in included, results may be considered to be accurate to +/- 10% of the patient true value.

Urine Creatinine: There is no change in urine creatinine standardization, units or reference intervals.

Creatinine Clearance: Measured creatinine clearance or calculated creatinine clearance using the Cockroft and Gault will change slightly with the change in serum the creatinine measurement. At serum creatinine concentrations below 200 umol/L, the creatinine clearance will be about 5-10 % higher than with the old assay, at higher creatinine concentrations the change is not significant.

Further Information:

More details are available on the following word document: Creatinine Assay Changes
or contact Dr Graham Jones:  8382-9160, gjones@stvincents.com.au