Measurement of urinary proteins over 24 hours is the definitive method to quantify proteinuria. However, prolonged collections of urine are inconvenient, cumbersome and often inaccurate due to frequent collection errors. To overcome these difficulties, this study estimated 24 hour urinary protein excretion from single voided, random urine sample. The protein-creatinine ratio of randomly obtained urine specimen correlated well with 24 hours urine protein with varying degree of proteinuria. This study was done on 70 patients having proteinuria from primarily glomerular diseases including diabetic and hypertensive kidney disease and excluded patients with orthostatic, benign, overflow and tubular proteinuria. This study showed that the random spot urine protein – creatinine ratio correlated well with the 24 hours measured urinary total protein as evidenced by the spearman’s correlation coeffecient. However the best positive correlation was seen in patients with non nephrotic proteinuria and less in patients with nephrotic range proteinuria.
As 24 h urine collections are cumbersome and not usually performed in the same 24 h of clinical assessment, an easier but reliable measure such as spot urine P/Cr would be ideal for usual clinical nephrology practice. Dipstick testing is known to be only moderately sensitive and specific for the detection of true proteinuria. A meta-analysis discussed by Craig et al. (80) of studies testing the threshold of urinalysis 1 + against the outcome of 24 h urine protein > 300 mg/day found sensitivities ranging from 67–100% and specificities 36–98%, although it should be noted that more than half the subjects in this analysis were pregnant, with or without hypertension.
Good correlations have been found between spot urine P/Cr and 24 h urine P/Cr in a large number of studies. All compared spot P/Cr with 24 h urine collections in the same 24 h period, some collecting morning urine samples. Although correlations were good in most of these studies very few tested agreement between methods. To advise replacing an established technique with a new one, it is first important to show that the two techniques agree sufficiently.
Table 1
Main studies comparing 24-hour Proteinuria with Protein/Creatinine ratio in spot urine
Study name
|
No of patients, Year
|
Study population
|
Main finding
|
Shaw AB, et. al. [10]
|
81,
1983
|
Patients with kidney disease and healthy individuals
|
P/C offers a good estimate of Prot/24, and it is indicated for screening in patients who probably do not adequately
collect 24-h urine
|
Villafruela JJ,et.al. [11]
|
452, 1990
|
Patients with kidney failurestratified according to eGFR
|
Correlation influenced by types of renal disease, degree of deterioration in kidney function and degree of proteinuria. Weak correlation in cases of severe kidney failure, interstitial nephritis or severe proteinuria.
|
Ruggenenti P, et al. [12]
|
98, 1998
|
Non diabetic nephropathies,
persistent proteinuria
|
First urine P/C is correlated with Prot/24 P/C has greater predictive power for
progression of kidney disease.
|
Price CP, et al.[13]
|
16, 2005
|
Pregnant women Kidney disease Kidney transplants From Rheumatology visits
|
High degree of correlation between P/C and Prot/24
P/C can be used for proteinuria screening
|
Antunes VV, et al. [14]
|
41, 2008
|
Glomerulardisease with:Nephrotic syndrome due to focal and segmental glomerulonephritis or membranous nephropathy, Non - nephrotic proteinuria
|
P/C is a precise method for determining proteinuria.
- The greater the degree of proteinuria (determined by the
Pearson correlation coefficient) the less the correlation between Prot/24 and P/C.
|
Guy M, et al.[15]
|
86, 2009
|
Diabeticnephropathy or Damage due to AHTor Renovasculardisease or Glomerulonephritis/ glomeruloesclerosis
|
Strong correlation between Prot/24 and P/C of 2nd and 3rd
sample.
A/C presents strong correlation with Prot/24 when this is > 1g/d.
|
Lambers Heerspink HJ, et al. [16]
|
701, 2010
|
Diabetic nephropathy, participating in RENAAL
|
A/C showed the strongest association with the risk of renal
events.
- Prot/24 has a weaker association; after calculation of P/C the
predictive value increases.
|
Methven S,etal. [17]
|
6842, 2010
|
Interstitial disease
Multisystemic disease
Diabetic nephropathy
Primary glomerulonephritis
Chronic kidney disease of unknown cause
|
P/C is more sensitive and more specific for the detection of proteinuria > 0.5-1g/d
Total proteinuria is not accurately determined by A/C due to the variable elimination of other type of proteins.
A/C and P/C vary substantially
according to age, gender, eGFR (effect related with muscle mass).
|
Methven S,etal. [17]
|
5586, 2011
|
> 18 years
Without beginning kidney replacementtherapy (including transplant)
Follow-up during > 1 year
|
Prot/24 predicts mortality and vascular events as equally as P/C and A/C. A/C and P/C can stratify cardiovascular risk in kidney disease patients equally.
- P/C is strongly correlated with low levels of Prot/24 (where it was traditionally considered that A/C
was better).
|
Methven S, et al. [18]
|
5586, 2011
|
> 18 years
Without beginning kidney replacementtherapy (including transplant)
Follow-up during > 1 year
|
ScreeningwithP/Cidentifies additional 16% of patients with significantproteinuria,not identified with A/C.
|
Similar study was also done by Yadav et al. in 2010 in Nepal which showed similar results in diabetic patients, as compared to this study [9]. He found an excellent sensitivity of 95.6% and specificity of 74.5% were achieved to detect proteinuria at the P: C ratio cutoff greater than 0.15. With this cutoff, the positive predictive value was 61.1% and negative predictive value was 97.6%. At a cutoff 2.43, the specificity was 100.0% however the sensitivity drops to 52.2%, whereas at cutoff 0.07 the sensitivity became 100.0% limiting specificity only to 5.4%.
Table 2
Correlation of Protein:Creatinine ratio with 24 hour urine protein excretion shown in previous studies
AUTHOR, YEAR
|
CORRELATION COEFFICIENT (r)
|
Ginsberg et al, 1983
|
0.97
|
Houser et al, 1984
|
0.98
|
|
0.97(ambulatory patient)
|
Schwab et al, 1987
|
0.98(hospitalized patient)
|
Lemann and Doumas, 1987
|
0.97
|
Boler et al, 1987
|
0.85(Normal pregnancy)
|
|
0.95(Twin pregnancy)
|
|
0.96(hypertensive pregnancy)
|
Ralston et al, 1988
|
0.92
|
Jachevatsky et al, 1990
|
0.92
|
Abitbol et al, 1990
|
0.95
|
Combs et al, 1991
|
0.98
|
Dyson et al, 1992
|
0.77
|
Mitchell et al, 1993
|
0.98
|
Quadri et al, 1994
|
0.92
|
Steinhauslin et al, 1995
|
0.93
|
Young et al, 1996
|
0.8
|
Robert et al, 1997
|
0.94
|
Sudan et al, 1997
|
0.93
|
Ramos et al, 1998
|
0.94
|
Evans et al, 2000
|
0.95
|
Rodriguez Thompson et al, 2001
|
0.8
|
Chitalia et al, 2001
|
0.97
|
Torng et al, 2001
|
0.79
|
Yamasit et al, 2004
|
0.95
|
Leung YY et al, 2006
|
0.91
|
Alfredo et al, 2007
|
0.98
|
Wahbeh et al, 2009
|
0.83
|
Though the result of this study agrees with most of the authors, but still there are some well known conflicting results. Some authors have reported only a medium correlation as shown in the table below. These discrepancies are probably due to an increase in tubular secretion of creatinine in various rates in patients with established renal disease and the interference of ketone bodies and some drugs in different creatinine assay method.
Table 3
Medium correlation shown between spot urine Protein:Creatinine ratio with 24 hour urine protein excretion shown in previous studies [9]
AUTHOR, YEAR
|
CORRELATION COEFFICIENT (r)
|
Lindow and Davey, 1992
|
0.53
|
Al et al, 2004
|
0.56
|
Durnwald and Mercer, 2003
|
0.64
|
Aggarwal N et al, 2008
|
0.596
|
Similarly various such studies have been carried out in India, comparing the spot urine PCR with the 24 hours UTP, recently in 2014 Patil et al. who evaluated 208 cases and 33 normal volunteers and found a strong correlation between the two with a Pearson‟s correlation coefficient of 0.9 and p < 0.05 [19].
Rahman et al. reported a significantly positive correlation (p < 0.05) in severe CKD patients in a study done in Bangabandhu Sheikh Mujib Medical University, Bangladesh on 50 non diabetic CKD patients in 2006, in this study there was a further very highly significant positive correlation among mild to moderate CKD patients [20].
Against this background of previous literature the K/DOQI Guidelines suggest measuring protein-to-creatinine ratio in spot urine samples, preferring a first morning specimen [21], whereas the Australian and New Zealand Society of Nephrology CARI Guidelines have made a suggestion for clinical care (based on level 3 and 4 evidence) that, "timed urine collections are recommended when knowledge of absolute level of proteinuria is required and that protein creatinine ratio (PCR) is an accurate test for diagnosis of proteinuria (> 1 g/day) but timed urines are required to establish a baseline" [22]. The present data support both these statements, demonstrating the potential for significant errors based on spot P/Cr estimation when 24 h protein excretion is high and the utility of spot P/Cr as a tool for assessing threshold protein excretion in population studies.
Some of the current clinical practice guidelines such as the KDOQI [21], CARI [23], KDIGO [24], UK Renal Association [25], NICE [26], and CSN [27] define the presence of proteinuria according to the P/C ratio (different value according to the Scientific Society). On the other hand, other guidelines such as theSEN-semFYC [28] and the ADA [29] define the appearance of proteinuria according to the urine albumin/creatinine ratio. Recently, a consensus document was published on the evaluation of proteinuria in the diagnosis and follow-up of chronic kidney disease patients that determined that in the detection and monitoring of proteinuria or albuminuria, it was not necessary to collect 24h urine [30]. In this study, given the difficulty of finding uniformity in the different guidelines and scientific societies for the initial determination of P/C or urine albumin/creatinine, this study focused on the analysis of the P/C ratio.
Implication of this study for clinical practice
If the desired outcome is to determine whether a patient has significant proteinuria (>1 g/day) but the absolute amount does not matter to the clinician or researcher then a spot urine P/Cr is an easy and reliable test, which will estimate 24 h proteinuria somewhere in this range. In contrast, if the clinician wishes to know the exact amount of protein excretion then a spot P/Cr will likely be within clinically acceptable limits only when proteinuria is at reasonably low levels. Some may argue that knowing the exact amount of protein excretion is unnecessary in routine clinical practice, but with the emphasis on reducing protein excretion in most renal disorders, some may disagree with this view. Furthermore, 24 h urine, although inconvenient, can offer the opportunity to assess urine volume and sodium excretion as an adjunct to management in some patients.
Limitations of this study
The main limitation is the need to increase the number of patients studied, since it was too small, specifically in the sub-group of patients who showed proteinuria within the nephrotic range. Another limitation is that no cause of kidney disease was excluded, including also a sample of patients with CKD, ESRD, toxemia of pregnancy and Glomerulonephritis. Increasing the number of samples collected and perhaps stratifying by underlying kidney pathology would help acquire a better knowledge of the correlation between the two techniques studied.
Third limitation may be due to the fact that no P:C ratio cut off value was used as an inclusion criteria for the patient. The adequacy of the 24 hours urine could not be ascertained due to lack of resources. In some cases, there also was a delay in obtaining the samples for 24 hours urinary protein analysis which might also have influenced the value.
Also as no blinding method was used, there might have been some selection bias. This study evaluated mainly patients suffering from glomerular diseases, all attempts were made to select the sample size properly however some patients may also be suffering from tubular disorders secondary to diabetes or other coexisting disorders. Evaluating properly for other coexistent conditions may have further influenced the correlation value.