Reliability of urinalysis for identification of proteinuria is reduced in the presence of other abnormalities including high specific gravity and hematuria

Purpose: Chronic kidney disease (CKD) is classified according to cause, glomerular filtration rate, and proteinuria. Identification of proteinuria with urinalysis (UA) is less accurate than quantification via other methods. We investigated factors leading to discordant UA findings when compared against paired albumin-to-creatinine ratio (ACR) testing. Methods: Four thousand three hundred and twenty-three UAs were grouped by proteinuria level (A1-A3); concordance with ACR was examined. Classification of UA with confounding factors (UA+CF) or without (UA-CF) was based on CF that resulted in >10% increase in false-positive proteinuria readings. The presence of >= 3+ blood, >= 3+ leukocyte esterase, any ketonuria, specific gravity >= 1.020, >= 1+ urobilinogen, >= 2+ bilirubin, >= 2+ bacteria, >= 3 RBC/hpf (high powered field), >= 10 WBC/hpf, and/or >= 6 epithelial cells/hpf led to UA+CF classification. Results: Proteinuria was determined to be present in 14.1% by UA dipstick and 24.9% by ACR. Using ACR as the standard, overall concordance was 80.4%, with 17.2% false-negatives and 2.3% false-positives by UA. UA+CF represented 55.6% of UA overall (n = 2404), and 98.0% of those false-positive for proteinuria. High specific gravity and hematuria are the strongest predictors of false positives. For A2 proteinuria (30- 300 mg/g, 1+,2+,3+ on UA) UA-CF had a higher negative predictive value (NPV) (99.8%) than UA+CF (77.6%); NPV for A3 proteinuria (> 300 mg/g, 4+ on UA) was 100% for UA-CF and UA+CF. Conclusion: Additional abnormalities were noted in > 50% of outpatient UAs indicating proteinuria. Given the significant proportion of patients having a false-positive UA for proteinuria when these CFs were present, we recommend that such patients undergo ACR confirmatory testing, according to a clinical algorithm for the incorporation of UA results into the management of CKD. (C) 2020 Elsevier Inc. All rights reserved.