Diagnostic accuracy of random urinary protein-to-creatinine ratio for proteinuria in patients with suspected pre-eclampsia.

Purpose: To evaluate the correlation between urine protein/creatinine ratio (UPCR) and proteinuria in a 24-h urine collection and to calculate the predicative accuracy of different cutoffs of UPCR for the diagnosis of proteinuria.

Methods: A retrospective cohort study including women who admitted for the evaluation for suspected preeclampsia (PET) beyond 20 weeks of gestation in a single tertiary center. Both UPCR test and quantification of proteinuria using 24-h urine collection were obtained during their index hospitalization no more than 48 h apart. Women with pre-existing diabetes mellitus, known renal disease or proteinuria prior to pregnancy or chronic hypertension were excluded. Predictive accuracy of UPCR for several cutoffs of proteinuria was evaluated. Multivariate logistic regression analysis was used to assess diagnostic accuracy of UPCR in sub-populations according to obstetrical characteristics.

Results: Overall 463 patients were included. Of them 316 (68.3%) have 24-h urine protein collection of ≥ 300 mg/day. Mean gestational age at evaluation was 34.0 ± 3.4 weeks. Median (and interquartile range) time interval between UPCR and 24-h urine collection was 1.8 (1.6-1.9) days. Sensitivity and specificity of UPCR of 0.3 for predicting proteinuria ≥ 300 mg/day were 90.1% and 63.3%, respectively. The corresponding values for difference proteinuria cutoffs: ≥ 1000 mg/day and 5000 mg/day were 98.4, 100% and 29.1, 36.0%, respectively. The optimal UPCR thresholds for 24-h urine protein collection of ≥ 300 mg/day,  ≥ 1000 mg/day and 5000 mg/day were 0.31, 0.70 and 2.49, respectively. The predictive accuracy of UPCR > 0.30 in predicting proteinuria was unaffected by demographic and obstetrical characteristics as maternal age, pre-pregnancy BMI, gestational age at examination, creatinine levels or by multiple gestation [adjusted OR 18.27 (95% CI 9.97-33.47)].

Conclusion: UPCR was strongly correlated with various cutoffs of proteinuria obtained by 24-h urine collection. UPCR cutoff varied depending on the specific measured outcome. This correlation was not affected by gestational age at examination.