Contemporary Test Performance of the Random Urine Protein-to-creatinine Ratio

Macie L Champion; David A Becker; Claire McIlwraith; Christina T Blanchard; Jeff M Szychowski; Dhong-Jin Kim; Victoria C Jauk; Lorie M Harper; Brian M Casey; Alan T Tita

doi:10.1055/a-1786-8847

Contemporary Test Performance of the Random Urine Protein-to-creatinine Ratio

Am J Perinatol. 2022 May 6. doi: 10.1055/a-1786-8847. Online ahead of print.

Authors

Macie L Champion^{1

2}, David A Becker^{1

2}, Claire McIlwraith^{1

2}, Christina T Blanchard^{1

2}, Jeff M Szychowski^{1

3}, Dhong-Jin Kim^{1

2}, Victoria C Jauk^{1

2}, Lorie M Harper^{1

2}, Brian M Casey^{1

2}, Alan T Tita^{1

2}

Affiliations

¹ Center for Women's Reproductive Health, University of Alabama at Birmingham, Birmingham, Alabama.
² Department of Obstetrics and Gynecology, University of Alabama at Birmingham, Birmingham, Alabama.
³ Department of Biostatistics, University of Alabama at Birmingham, Birmingham, Alabama.

PMID: 35240698
DOI: 10.1055/a-1786-8847

Abstract

Objective: The random urine protein-to-creatinine ratio (UPCR) is a screening test used for predicting clinically significant proteinuria (urine protein ≥ 300 mg) during pregnancy. No consensus exists on the optimal random UPCR cutoff for performing follow-up 24 hour urine (24H) total protein collection. We aim to evaluate the test performance of random UPCR in predicting proteinuria in a contemporary cohort.

Study design: This was a retrospective cohort study of pregnant patients at our institution from 2014 to 2018 with a random UPCR and follow-up 24H protein collection. The primary analysis estimated the test characteristics (sensitivity, specificity, positive and negative predictive values) of using random UPCR for the detection of proteinuria defined as urine protein ≥300 mg on 24H protein collection. UPCR cutoffs from 0.10 to 0.30 mg/dL were evaluated, receiver operator characteristic (ROC) curve was constructed, and area under the curve (AUC) was determined. A secondary analysis examined the correlation between UPCR and 24H protein using least squares regression and Pearson correlation.

Results: Paired UPCR and 24H collection results were available for 1,120 patients. Mean gestational age at time of UPCR was 31.1 ± 5.1 weeks and 687 (61.3%) of patients had a 24H ≥300 mg. UPCR <0.10 mg/dL effectively excluded proteinuria ≥300 mg on 24H collection, while UPCR ≥0.18 mg/dL correctly classifies proteinuria with 91% sensitivity, 57% specificity, 77% positive predictive value, and 79% negative predictive value. UPCR ≥1.07 mg/dL had 100% specificity for 24 hour proteinuria. The area under ROC curve was 0.86. UPCR and 24H collection were highly correlated with a Pearson correlation coefficient of 0.85. After our institution lowered the threshold to obtain a 24H from UPCR ≥0.20 mg/dL to ≥0.10 mg/dL in May 2017, the percentage of patients meeting criteria for 24H collection increased from 57.8 to 84.4%.

Conclusion: The AUC and Pearson correlation suggest random UPCR is a high performance test for the prediction of proteinuria on 24H. Optimal test performance is dependent upon clinical consideration and upon the implications of the disease or condition. A random UPCR screen positive threshold of 0.18 mg/dL maximizes sensitivity to identify clinically significant proteinuria.

Key points: · Random urine protein to creatinine ratio is a high performance test for proteinuria.. · A random UPCR threshold of 0.18 mg/dL maximizes sensitivity to identify proteinuria.. · Optimal test performance is dependent on the disease or clinical condition..