KDOQI (Kidney Disease Outcomes Quality Initiative)


Executive Summaries of 2000 Updates | Anemia | Hemodialysis | Peritoneal Dialysis
Vascular Access | Nutrition | Chronic Kidney Disease | History of DOQI


KDOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification



Normal individuals usually excrete very small amounts of protein in the urine. Persistently increased protein excretion is usually a marker of kidney damage. The excretion of specific types of protein, such as albumin or low molecular weight globulins, depends on the type of kidney disease that is present. Increased excretion of albumin is a sensitive marker for chronic kidney disease due to diabetes, glomerular disease, and hypertension. Increased excretion of low molecular weight globulins is a sensitive marker for some types of tubulointerstitial disease. In this guideline, the term “proteinuria” refers to increased urinary excretion of albumin, other specific proteins, or total protein; “albuminuria” refers specifically to increased urinary excretion of albumin. “Microalbuminuria” refers to albumin excretion above the normal range but below the level of detection by tests for total protein. Guidelines for detection and monitoring of proteinuria in adults and children differ because of differences in the prevalence and type of chronic kidney disease.

Guidelines for Adults and Children

Specific Guidelines for Adults

Specific Guidelines for Children Without Diabetes

Specific Guidelines for Children With Diabetes


The measurement of urinary protein excretion provides a sensitive marker of many types of kidney disease from early to advanced stages. The most pertinent question with respect to screening for proteinuria is whether early detection of kidney disease associated with this abnormality will result in a more timely introduction of therapy that may slow the course of disease? The answer is “yes”—at least for some chronic kidney diseases. For example, in diabetic kidney disease, early detection of albuminuria appears to permit effective therapy early in the course of disease.

The American Diabetes Association8 and the NKF-PARADE6,7 have recommended assessment of proteinuria to detect chronic kidney disease. These recommendations largely agree in the methods for assessment of proteinuria. The purpose of this guideline is to review the rationale for methods of assessment of proteinuria and to determine whether detection and monitoring of proteinuria using untimed (“spot”) urine samples is as accurate as using timed (overnight or 24-hour) urine specimens. Algorithms for screening and evaluation of proteinuria in asymptomatic, healthy individuals and in patients at increased risk for chronic kidney disease recommended by NKF-PARADE are given in Part 9.


Criterion Standard

It is important to consider the timing of urine specimens and the methods for detection of urine proteins. Although the basic concepts of measuring and interpreting urinary protein excretion have changed little over several decades, clinicians must now decide whether simple qualitative or more cumbersome quantitative tests are necessary and whether albumin or total protein should be measured. In clinical practice, most screening methods use a commercial dipstick, which measures total protein or albumin. These dipsticks, which are of course simple to use, usually afford high specificity; ie, they have relatively few false positive results, thereby creating a practical advantage for the clinician. However, they afford low sensitivity; ie, they may fail to detect some forms of kidney disease during the early stages, when the level of proteinuria is below the sensitivity of the test strip used.

Timed urine collections versus untimed (“spot”) urine samples. When screening tests are positive, measurement of protein excretion in a 24-hour collection has been the longstanding “gold standard” for the quantitative evaluation of proteinuria. However, in recent years some studies have advocated that the measurement of protein excretion should be done on an overnight specimen. The rationale for measuring proteinuria in timed overnight urine collections rather than 24-hour specimens relates to the lack of consistency when hourly protein excretion rates are examined in the same individual at different times during the day. This inconsistency results from varying levels of activity and possibly other factors that are not well documented. The high intra-individual variability that ensues makes serial comparisons in individual patients very difficult unless multiple measurements are taken. This problem is particularly troublesome for individuals with orthostatic proteinuria—who may excrete more than 1 g of protein during waking hours, but less than 100 mg during sleep. Indeed, evaluation for postural (orthostatic) proteinuria requires comparison of a measurement of protein excretion in an overnight (“recumbent”) collection to a daytime (“upright”) collection.

An alternative method for quantitative evaluation of proteinuria is measurement of the ratio of protein or albumin to creatinine in an untimed “spot” urine specimen. These ratios correct for variations in urinary concentration due to hydration and provide a more convenient method of assessing protein and albumin excretion than that involved with timed urine collections. The issue to be explored in this section is whether this increased level of convenience can be achieved without a reduced level of precision. Based on the review of evidence accumulated over three decades, the Work Group proposes that the time has come to forego the traditional “timed urine collections” and adopt the use of “spot” urine measurements that compare the concentration of protein to the concentration of creatinine.

Total protein versus albumin. The assessment of protein excretion in the urine can be accomplished by several different techniques. In addition to standard methods of measuring total protein, there are now multiple versions of immunoassays capable of detecting albumin levels at concentrations present in the majority of normal people. In general, the literature does not provide substantial information concerning the relative merits of measuring total protein versus albumin to detect and monitor kidney damage. Different guidelines for children and adults reflect differences in the prevalence of specific types of chronic kidney disease.

Strength of Evidence

Rationale for ADULTS and CHILDREN

This section will describe the rationale for using “spot” urine samples to estimate protein excretion individuals of all ages, timing of urine samples, and dipstick for quantitative assessment. The two subsequent sections will review data for adults and children separately.

Rationale for Using “Spot” Urine Samples

Collection of a timed urine sample is inconvenient and may be associated with errors (R, O).  Twenty-four–hour urine collections may be associated with significant collection errors, largely due to improper timing and missed samples, leading to over-collections and under-collections. Timed overnight collections or shorter timed daytime collections may reduce the inconvenience of a 24-hour collection, but are still associated with collection errors. In addition, errors due to incomplete bladder emptying are relatively more important in shorter collection intervals.

Concentration of protein in a spot urine sample provides a rough index of the protein excretion rate, but is also affected by hydration (R, C). The concentration of protein in the urine is affected by urine volume as well as protein excretion rate. Urine volume is dependent primarily on the state of hydration. For example, in a patient with urine protein excretion of 500 mg per day the protein concentration may vary from 100 mg/dL (2+ on the dipstick) in a patient with urine volume of 500 mL/d to 20 mg/dL (trace on the dipstick) in a patient with urine volume of 2500 mL/day. Despite this, there is a rough correlation between protein concentration in a spot urine sample and protein excretion rate (Tables 53, 54, and 55).

kck_t53.gif (13964 bytes)

kck_t54.gif (37239 bytes)

kck_t55.gif (27048 bytes)

Urine protein-to-creatinine and albumin-to-creatinine ratios provide accurate estimates of the urinary protein and albumin excretion rate, and are not affected by hydration (R, C). Several studies have addressed the relationships between total excretion of protein or albumin and the ratio of either to creatinine in patients of all ages (Tables 56, 57, 58, and 59).

Since urine proteins and creatinine are highly soluble in water, they will undergo similar, if not identical, dilution in urine. In principle, if the excretion of creatinine is relatively constant throughout the day, and similar among individuals, then the ratio of protein-to-creatinine in an untimed sample would reflect the excretion of protein. Although creatinine excretion varies among individuals according to age, gender, race, and body size, the results from these studies in adults and children demonstrate a strong correlation between these measures.

Rationale for Timing of Sample Collection

A first morning urine specimen is preferred, but random urine specimens are acceptable if first morning urine specimens are not available (R, O). A first morning urine specimen is preferred because it correlates best with 24-hour protein excretion and is required for the diagnosis of orthostatic proteinuria. In children, orthostatic proteinuria must be excluded by a first morning urine protein measurement if the initial finding of proteinuria was obtained on a random specimen during the day. Otherwise, for ease and consistency of collection, a random urine specimen for protein or albumin to creatinine ratio is acceptable if a first-morning urine specimen is not available. This recommendation is consistent with the recommendations by the American Diabetes Association8 and by the NKF PARADE,6,7 which recommend a first-morning sample, but accept a random sample if a first-morning specimen is not available.

Table 60 compares the advantages and disadvantages of the various modalities of collecting urine for evaluating kidney function. For all procedures, identical methods are utilized in the laboratory. The differences among these protocols balance ease of collection of samples with the need to collect urine to reflect kidney function over the course of the day or overnight.

Rationale for Measurement Methods

Screening for proteinuria with urine dipsticks is acceptable. Confirmation of proteinuria should be performed using quantitative measurements (R, O). Standard urine dipsticks detect total protein above a concentration of 10 to 20 mg/dL. The reagent pad contains a colorimetric pH indicator dye which changes color when bound by negatively charged serum proteins, including albumin and most globulins. The standard urine dipstick is insensitive for low concentrations of albumin that may occur in patients with microalbuminuria. In addition, the standard dipstick is also insensitive to positively charged serum proteins, such as some immunoglobulin light chains.

Albumin-specific dipsticks detect albumin above a concentration of 3 to 4 mg/dL and are useful for detection of microalbuminuria.

Consistent with recommendations by ADA and NKF-PARADE, the Work Group recommended screening using either standard or albumin-specific dipsticks, or protein-to-creatinine or albumin-to-creatinine ratio. Screening with a dipstick for proteinuria or albuminuria is often a satisfactory first approach to evaluation of kidney disease; however, clinicians need to be cognizant of causes of false positive and more importantly false negative results (Table 61), and in both instances repeat analyses of urine with quantitative total protein or albumin and creatinine analyses are strongly advised when a result may be inconsistent with the clinical evaluation.

Special care should be taken to avoid false negative results which may delay implementation of treatment early in the course of kidney disease.

Monitoring proteinuria in patients with chronic kidney disease should be performed using quantitative measurements (O). Changes in proteinuria provide important prognostic information. Increasing proteinuria is associated with a higher risk of loss of kidney function. Decreasing proteinuria, either spontaneously or after treatment, is associated with a lower risk of loss of kidney function. Quantitative measurements provide a more accurate assessment of changes in proteinuria.

Rationale for Type of Protein: ADULTS

In adults, it is preferable to assess proteinuria as albumin, because:

In adults, the most common types of chronic kidney disease are due to diabetes, hypertension, and glomerular diseases. In patients with diabetes mellitus, there has been nearly a uniform adoption of albumin as the “criterion standard” in evaluating kidney damage. Thus, for this disease the same standards have been adopted for adults and children. Preliminary data suggest that elevated albumin excretion is also a marker of kidney damage in adults with hypertension. Proteinuria in glomerular diseases is primarily due to increased albumin excretion. Therefore, the Work Group concluded that albumin should be measured to detect and monitor kidney damage in adults.

The interpretation of albuminuria in kidney transplant recipients is more complicated than in other patients with chronic kidney disease. First, depending on the interval since transplantation, the patients’ native kidneys may still excrete small amounts of protein, which may be sufficient to cause a positive test for albumin. Second, the main causes of damage in kidney transplant, rejection or toxicity from immunosuppressive drugs, are not characterized by proteinuria. However, diabetic kidney disease is the underlying cause for a large fraction of kidney transplant patients, which may recur in the transplant. Moreover, hypertension is very common after transplantation and is strongly associated with a more rapid loss of kidney function in transplant patients. Finally, recurrent glomerular disease may occur after transplantation and is associated with a greater risk of graft loss. Albuminuria is a better marker than total urine protein of kidney damage due to diabetes, hypertension, and glomerular disease. For these reasons, the Work Group recommends testing and monitoring for albuminuria, rather than total protein, in kidney transplant recipients, as well as in patients with other causes of chronic kidney disease.

The cost or technical difficulty of measuring albumin may exceed that for measuring total protein. It is acceptable to measure total protein-to-creatinine ratio as an index of proteinuria in adults when albumin-to-creatinine ratio is substantially elevated (eg, >500 to 1,000 mg/g). However, there is no reliable method to convert ratios of albumin-to-creatinine to total protein-to-creatinine or vice versa.

Rationale for Type of Protein: CHILDREN WITHOUT DIABETES

In children without diabetes, it is preferable to assess proteinuria as total protein, because:

The prevalence of chronic kidney damage due to diabetes and hypertension is far lower in children than in adults. In contrast, the prevalence of kidney disease due to urinary tract abnormalities and congenital tubular disorders is far more common in children than in adults.219 These latter diseases may be characterized by low molecular weight proteinuria, which would be detected by tests for total urine protein, but not by tests for albumin. Therefore, the Work Group recommends that total urine protein should be measured to detect and monitor kidney damage in most children, one exception being children with diabetes mellitus.

Rationale for Type of Protein: CHILDREN WITH DIABETES

In post-pubertal children with duration of diabetes greater than 5 years, it is preferable to assess proteinuria as albumin because:

In other children with diabetes, it is preferable to assess proteinuria as total protein because:

The risk of diabetic kidney disease in children is higher in post-pubertal children with duration of diabetes greater than 5 years than in other diabetic children. For these reasons, the American Diabetes Association recommends screening these children for chronic kidney disease, using the same algorithm as for adults. Other diabetic children are screened using the same algorithms as for other children.


The main limitations of assessment of proteinuria as a marker of chronic kidney disease is potential misclassification of individuals due variability of levels of total protein or albumin in an individual over time and the extent to which conditions at the time of testing may obscure the true level. Excretion of total protein or albumin in the urine are highly variable in individuals with or without kidney disease. Most studies suggest that the standard deviation is about 40% to 50% of the mean. Examples of conditions that affect protein excretion other than kidney disease include activity, urinary tract infection, diet, and menstruation. Attempts to avoid these pitfalls include careful definition of events that should preclude the interpretation of abnormal results and consideration of repeat studies when abnormal results are obtained. Some authors have advocated that multiple (up to 5) specimens be obtained in order to obtain a reliable result.42 The Work Group does not believe that such an approach is feasible in most instances. However, the Work Group acknowledges the need to repeat abnormal tests, especially low levels of total protein or albumin and the necessity to carefully consider the clinical setting in interpretation of urine protein measurements.

A limitation of this guideline is the use of correlation coefficients, rather than more detailed assessments of precision and bias, to assess the accuracy of spot urine measurements of protein-to-creatinine ratios as a measure of protein excretion rates. In most cases, data were not available to characterize the precision and bias. In addition, other than distinguishing normal from abnormal, the exact level of proteinuria is not usually required for clinical decision-making. In most circumstances, a “rough” approximation of the level of proteinuria or changes over time are sufficient. Thus, the Work Group concludes that the uniformly high correlation coefficients are sufficiently strong evidence to warrant the conclusions presented here.

Clinical Applications

General considerations for adults and children

The identification of persistent proteinuria or albuminuria in patients of all ages has importance when considering diagnosis, prognosis, and therapeutic options. The relative ease with which proteinuria can be assessed and monitored allows clinicians to identify individuals with completely asymptomatic forms of progressive kidney disease during the early stages of their disease. Such patients may benefit from subsequent changes in management that forestall or prevent additional kidney problems.

Proteinuria is a key finding in the differential diagnosis of chronic kidney disease. The relationship between the level of proteinuria and the type (diagnosis) of chronic kidney disease is reviewed in Guideline 6 and in Part 9.

Proteinuria is a key prognostic finding in chronic kidney disease. The prognosis of patients with a variety of kidney disorders often correlates with their level of and persistence of proteinuria over time—even when other variables are controlled. This is important because of the obvious therapeutic implications for patients who are in the high risk category that is characterized by persistent, heavy proteinuria. The relationship between the level of proteinuria and risk for loss of kidney function is considered further in Guideline 13.

Finally, the most important clinical application of defining patients with proteinuria is potentially beneficial therapy. Many lines of evidence now indicate that medications that reduce proteinuria may provide significant long term benefits for patients with chronic kidney disease.

Specific considerations for children

The optimal frequency and timing of urine screening for proteinuria in children have not been well established. At one end of the spectrum, the governments of some countries have mandated that such screening be done on all school children every year. In Japan, for example, this has been in place for over 25 years.220 In contrast, the American Academy of Pediatrics recommends that such screening be done on two occasions during childhood—once before starting school and then again during adolescence.221 Some authors even consider this to be excessive and have proposed that urinalysis screening be limited to a single first morning dipstick done at school entry age in all asymptomatic children.222

The recommendations described in this section for children are consistent with other recent publications from the National Kidney Foundation PARADE6,7 and the American Diabetes Association.223 It should be noted, however, that the ADA position statement draws attention to the fact that microvascular disease may occur in prepubertal children and that “clinical judgment should be exercised when individualizing these recommendations.”

Implementation Issues

The implementation of the guidelines in this section will encounter at least two potential obstacles. The first is the widely held belief that 24-hour urine collections provide “the only accurate method” of measuring protein or albumin excretion. This even applies to some pediatricians who continue to request 24-hour urine studies in small children despite the high degree of difficulty involved.

The second potential problem involves the adoption of urine protein measurements factored by urine creatinine. This approach has been developed to some extent for urine calcium-to-creatinine measurements, but many physicians are not aware of the accuracy and validity of protein-to-creatinine ratios. Many clinical laboratories may not report ratios of analytes to creatinine. A significant amount of education will be necessary to implement this approach.

A less obvious implementation issue relates to measuring albumin rather than total protein in the urine specimens. Assays for albumin may not be as available as those for total protein in some smaller communities. In such instances, the use of a spot urine and expression of the urine protein-to-creatinine ratio is still preferable to the 24-hour collection.

Research Recommendations

Evaluate novel approaches to measuring urine and blood abnormalities which may predate and possibly predict proteinuria/albuminuria. Examples include elevated levels of 2-microglobulin and other tubular proteins in the urine of diabetic patients. Additional efforts should be instituted to identify constituents present in blood and/or urine that indicate normal kidney function with high specificity.

It would be useful to conduct prospective trials of the long-term efficacy of antihypertensive medications that reduce albumin/protein excretion in kidney disease. These studies should incorporate better procedures to examine the efficacy of sustaining kidney function in advanced kidney disease and in reducing the incidence of cardiovascular disease in patients with kidney disease.224-227

It would also be useful to determine the relationships between factors that may affect albumin/protein excretion and also increase the risk of macrovascular disease (eg, glucose intolerance/diabetes mellitus, rising blood pressure, elevated lipid levels, and obesity) and progressive kidney failure.228,229

2002 National Kidney Foundation, Inc