How many urine samples are needed to diagnose and monitor albuminuria?
Urinary albumin excretion is a powerful predictor of cardiorenal events, but has high test-test variability. This problem can be addressed by basing clinical decisions on results from more than one sample. However, one of the challenges in determining the appropriate number of samples to test is that there are very few studies that have quantified variability in the most widely used measurement of albumin excretion — the urine albumin:creatinine ratio (UACR). Doing so allows the development of tools that can be used to guide clinical practice in regard to diagnosing albuminuria and determining how likely it is that any observed change in UACR over time reflects a true change. In the Progression of Diabetic Complications (PREDICT) cohort study, 826 Australians with type 2 diabetes (mean age 66+/-9y, 65% men) provided four random spot UACR samples within a 28-day period. These results, published in the American Journal of Kidney Diseases, were used to generate the following clinical tools that provide an evidence-based, practical approach to diagnosing and monitoring albuminuria.
Determining the number of urine samples needed to monitor albuminuria over time
We have developed an online calculator to determine the probability that a change in urine albumin:creatinine ratio (UACR) between two timepoints (e.g. before vs. after treatment) has exceeded a minimum clinically meaningful difference.
Launch the Urine ACR change probability calculator
Determining the number of urine samples needed to diagnose albuminuria
For diagnosis of albuminuria in clinical settings, high test-test variability is usually addressed by basing diagnoses on at least two samples. However, this suggests that two samples might also be necessary to confidently exclude albuminuria. Furthermore, since the science behind the precise number of samples is not strong, it is possible that more samples might sometimes be necessary. Using the geometric mean of four samples as the reference standard, the ability of means based on one, two and three samples to predict the four-sample classification into albuminuria or no albuminuria was determined.
This analysis was used to produce a decision tree, which, depending on the level of UACR, determines whether or not further samples are required to confidently classify an individual. When the first-sample value or the geometric mean of two or three samples lies within the uncertain range, a further sample is required for accurate classification. The decision tree is based on the geometric mean, rather than the simple arithmetic mean, of the UACR. Geometric mean can be calculated using our Urine ACR Change Probability Calculator by entering UACR values in either timepoint, clicking ‘Calculate’ and checking the Urine ACR Summary table.
Decision tree showing the number of urine collections required for UACR classification, according to UACR values.
Click image to enlarge
Albuminuria diagnosed based on KDIGO guidelines as ≥3.0 mg/mmol (2).
References
- Rasaratnam N. et al. Within-individual variability of the urine albumin/creatinine ratio in people with type 2 diabetes: clinical and research implications. American Journal of Kidney Disease. 2024.
- Rossing P. et al. Executive summary of the KDIGO 2022 Clinical Practice Guideline for Diabetes Management in Chronic Kidney Disease: an update based on rapidly emerging new evidence. Kidney International (2022) 102, 990–999.