Many of the points raised by Drs. Mogensen and Korosi are discussed in the full version of the guidelines (1). Page numbers cited below refer to pages in that publication.

Dr. Mogensen agrees with the recommendation to test patients at increased risk for chronic kidney disease for "microalbuminuria" in a spot urine sample. However, he suggests measuring the albumin-to-creatinine ratio rather than an albumin-specific dipstick, as depicted in Figure 2 of our article in Annals. We agree that that testing could begin with measurement of albumin-to-creatinine ratio (page S215), as discussed in a more recent consensus conference sponsored by the National Kidney Foundation and the National Institute of Diabetes and Digestive and Kidney Diseases (2). The gold standard method for detection of microalbuminuria is based on immunoassay in a timed urine collection. Sensitivity of detection of microalbuminuria using an albumin-specific dipstick or an albumin-to-creatinine ratio in spot urine samples is 80% to 90% (pages S97–S98). The guidelines also recommend periodic reevaluation of patients with negative test results using either method.

Currently available methods are sensitive enough to detect urine albumin concentrations just above the normal range (Table) (3). Ultimately, many factors, including cost, influence the decision about whether testing should begin with a dipstick in the physician's office or with a laboratory test. Clinicians must also be attentive to common causes of false-positive and false-negative results (page S99). A recent report demonstrated limited sensitivity of an immunoassay compared with high-performance liquid chromatography (4). More studies are needed to determine the appropriate reference range and clinical importance of albuminuria detected by high-performance liquid chromatography.

Dr. Korosi agrees with the recommendation to estimate GFR from serum creatinine measurements using Dr. Mogensen agrees prediction equations but also advocates collection of 24-hour urine samples for confirmation of GFR and estimation of dietary protein intake. Estimating GFR from the mean of the 24-hour urea and creatinine clearance has been validated only in individuals with GFRs less than approximately 15 mL/min per 1.73 m² (5). Studies in patients with higher GFRs show that 24-hour creatinine clearance does not provide a more accurate estimate of GFR than the Modification of Diet in Renal Disease Study prediction equation. Thus, the guidelines do not recommend routine collection of a 24-hour urine sample to confirm the estimate of GFR in routine practice. In subspecialty practice, a 24-hour urine sample can be useful to confirm GFR estimates lower than 15 mL/min per 1.73 m² and to estimate dietary protein intake (pages S89–S90). Other indications for clearance measurements to estimate GFR (S90) and recommended filtration markers (S77) are discussed in the guidelines.