Abstract BACKGROUND AND AIMS It has been postulated that the protein-creatinine ratio in spot urine sample (uPCR) could be an adequate alternative to measurement of proteinuria in 24-h urine samples (24-h proteinuria), which despite being the gold standard for the diagnosis and follow-up of patients with kidney disease, its accuracy is frequently influenced by incorrect sample collection. Several studies have shown a correlation between uPCR in spot urine samples and 24-h proteinuria. However, in daily clinical practice we often find significant differences between both measurements. Therefore, we set out to conduct this study to analyze the concordance between uPCR in first morning spot urine samples and 24-h proteinuria in patients with biopsy-proven glomerular diseases. METHOD We identified patients with biopsy-proven glomerular diseases who were evaluated during their disease course at our center between January 1, 2005 and December 31, 2021, and included patients with paired 24-h protein excretion measurement and spot urine samples that measured protein and creatinine concentration, simultaneously collected at the same day. We excluded patients with interstitial or vascular causes for kidney disease. We performed correlation analyses of uPCR and 24-h proteinuria globally and analyzed the influence of demographic data as well as eGFR and type of glomerular disease in this relationship. The concordance between both measurements was compared using intra-class correlation coefficients (ICC) and Bland–Altman plots. RESULTS About 351 patients with simultaneous uPCR and 24-h proteinuria measured at the same day were included in the study, 67.2% were male and the mean age was 53.9 ± 17.9. Median serum creatinine was 1.28 mg/dl (IQR 0.95–1.86) and mean eGFR measured by CKD-EPI was 60.1 ± 30.6 mL/mi/1.73 m2. Median 24-h proteinuria was 1.70 g/day (IQR: 0.62–3.45) and median uPCR was 1.09 g/g (IQR: 0.35–2.70). Wilcoxon rank test showed a systematic underestimation of uPCR compared with 24-h proteinuria (Z = −8.22, P < 0.001). Mean ICC for both measurements was 0.82 (CI 95%: 0.77–0.85, P < 0.001). There was a significant agreement between both measurements after stratifying according to eGFR, with a mean ICC in patients with GFR >60 mL/min of 0.79 (CI 95%: 0.71–0.85, P < 0.001), mean ICC in patients with GFR between 30 and 60 mL/min of 0.80 (CI 95%: 0.72–0.86) and a mean ICC of 0.85 (CI 95%: 0.74–0.91, P < 0.001) in patients with GFR < 60 mL/min. Agreement between Upcr and 24-h proteinuria did not vary according to age, sex or type of glomerular disease. After dividing the cohort into three categories of proteinuria (<1, 1–3, >3 g/day for 24-h proteinuria, and < 1, 1–3, >3 g/g for uPCR), we performed Cohen's κ test to determine agreement, finding moderate agreement between measurements (κ = 0.46, P < 0.001). Passing-Bablok Regression and Bland-Altman plots indicated high concordance between the two methods when proteinuria <3.5 g/day, while outliers in the data suggested greater variability between measurements for patients with proteinuria over 3.5 g/day (Figure 2). CONCLUSION The level of concordance between proteinuria from 24-h urine samples and protein–creatinine ratio from spot urine samples is high regardless of eGFR, age, sex and type of glomerular disease. However, the level of agreement decreases in nephrotic-range proteinuria. Therefore, we suggest that uPCR could be an appropriate measure to assess and follow-up patients with glomerular diseases and non-nephrotic range proteinuria, while the degree of proteinuria should still be monitored with 24-h urine specimens in patients with proteinuria >3.5 g/day.
Read full abstract